Cyprotex Discovery Ltd. v University of Sheffield

Background to the Dispute

This case is concerned with the ownership of the copyright in a set of computer programs arising out of research work undertaken at the University of Sheffield (“Sheffield”) by the Molecular Pharmacology and Pharmogenetics Section of the Department of Clinical Pharmacology which a programmer employed by Cyprotex Discovery Limited (“Cyprotex”) subsequently developed into a potentially commercially exploitable form. The case highlights the desirability of parties involved in computer program development, before they start development work on any program, to provide clearly in their contract for who is to own the copyright and exploitation rights of the development work at each stage of a program’s development. The law of copyright is still adapting to the intricacies of information technology and it sometimes fails to provide a clear answer to problems that can arise as to the ownership of copyright and exploitation rights in such development work.

The dispute is concerned with the Simcyp Project and the starting point is research work that was commenced by Professor Tucker in the mid 1980s. When Dr Rostami joined Professor Tucker’s Section of the Department in 1992, he became an active collaborator in this work. The Simcyp project is concerned with pioneering a computer-based aid to assist pharmaceutical research and development in seeking to predict how a potential new drug would be absorbed across the gut wall and other barriers within the body, distributed to the various organs and tissues of the body and eliminated from the body by excretion and metabolism. The Simcyp project aims to provide drug companies with a rapid and reliable means of evaluating these potential characteristics of such a drug. Given the huge and often abortive costs of drug research and development, a successful and reliable way of providing such predictions has a significant commercial potential.

The Simcyp Project involved the creation of a set of algorithms and databases using a software modelling system in the form of MathCAD worksheets. In that form, Simcyp could not be widely used and there was a need to develop the Simcyp algorithms and databases from the worksheets into a program with a user-friendly and widely useable operating environment such as Windows. The development of such a program required finance since it would need skilled programming expertise and the development of many additional databases.

As a result, Sheffield decided to raise the finance from drug companies who would each sponsor the development by subscribing a pre-determined share of the projected cost and the development of the program would proceed in two phases. The first phase would involve the production of a user-friendly program using a retained programmer. The sponsors would have limited rights to evaluate the program and suggest changes during this development phase. The second phase would involve the development of databases required by the program which would then be available for commercial marketing. The sponsors would have obtained input into the program’s facilities and would be granted a license on favourable terms to use the program following its successful development.

Cyprotex was formed in May 2001 as a management buyout of an operating division of Medeval Ltd, a contract research company. The relevant division was called Cyprotex and this was devoted to the development of computer programs for use in drug behaviour predictions and test trial simulations. Professor Tucker sat on the Scientific Advisory Board of Cyprotex and, as a result of this contact, Cyprotex entered into a contractual arrangement with Sheffield to provide the necessary computer programming expertise. The necessary payment would be made by Cyprotex being offered sponsorship rights in the project which would enable it to obtain a valuable license to use the program on favourable terms. The principal programmer who undertook most of the programming work was Dr Edwards.

The sponsors provided their share of the development costs through a multi-party Research Agreement in which each sponsor and the University were parties. The programming services provided by Cyprotex and Cyprotex’s rights in the software produced by Dr Edwards were provided by and defined in this same Research Agreement. All went well to begin with and between June 2000 and March 2001. Dr Edwards produced a first full version of the program and had started work on a second full version following the agreement with Dr Rostami of new requirements and the production of a new specification. The Research Agreement itself is undated but took effect when the last signature was added to it on 21 December 2001. There is an issue as to whether, and if so in what way, this Agreement had retrospective effect with regard to Dr Edwards’ programming work since it defined the period of performance as starting on 1 March 2000 and Dr Edwards’ work had started in June 2000. There is also an issue as to the extent to which Cyprotex may take advantage of the Research Agreement since it was not incorporated until 31 May 2001.

Unfortunately, by September 2001 with Dr Edwards still working on the second version of the software, the parties were in dispute and their commercial relationship broke down. By then, Sheffield had formed a company called Simcyp Ltd with the aim of using it to market and exploit the Simcyp algorithms, software and databases. This company asserted that Sheffield owned the copyright in the software being developed by Dr Edwards, an assertion stoutly denied by Cyprotex who made equal and competing assertions as to ownership of the copyright. Dr Edwards was then instructed by Cyprotex to cease development work.

The dispute originated during negotiations in the summer of 2001 that had been initiated by Cyprotex, relating to the commercial exploitation of Simcyp and its wish to acquire from Sheffield such remaining rights in Simcyp that it did not then own. During these negotiations, it became clear that Sheffield was maintaining that it held the copyright in the Simcyp programs developed and being developed by Dr Edwards, in part by relying on provisions in the Research Agreement so that any separate use or licensing of Simcyp by Cyprotex would amount to an infringement of Sheffield’s copyright. Cyprotex, on the other hand, maintained that it had exclusive rights over Dr Edwards’ development and programming work.

Following the breakdown of negotiations, Cyprotex started these proceedings claiming a declaration that it was the owner of all copyright and rights of exploitation in the Java Class Libraries and Graphical User Interface programs, being the technical description of Dr Edwards’ development work. Sheffield’s principal relief claimed by counterclaim is a claim for a declaration that it was the owner of similar rights, albeit that the rights are expressed as being in relation “to or over all and any works relating to Simcyp created by Duncan Edwards”. However, additionally, it claims an entitlement to a copy of the source code used to develop the Libraries and Interface programs even if Cyprotex is found to be the owner of the intellectual property and commercial exploitation rights in these programs. This right is claimed as a result of what Sheffield alleges to be the terms of an on-screen license that the parties had agreed to be the basis of Cyprotex’s entitlement to use Simcyp.

The action was started in the Chancery Division in the High Court in Leeds on 27 September 2001 and hearings were held before Judge Behrens on 21 December 2001 and 10 and 17 January 2002. A number of orders were made by consent involving the delivery up of the software in issue, disclosure of the source code, nondisclosure to third parties and non-use pending trial, security for costs and for the non-use of the program pending trial. On 20 March 2002, following an application to the TCC in London by the claimant, the action was ordered to be transferred to the TCC, it was rota’d to me and I have given detailed case management directions. The trial was held in September 2002.

In order to resolve these competing claims, I must first decide whether the entirety of the parties’ relationship is governed by the Research Agreement or whether at least part of the development work was undertaken pursuant to an earlier informal agreement between the parties. I must then determine what express and implied terms are applicable to the parties’ relationship and construe these relevant provisions against the common factual background known to both parties at the time of entering into any such contract, applying any appropriate meaning or gloss to relevant contractual provisions that arises from their use in their highly technical context and taking into account any applicable provision of the law of copyright. Only then can I determine what declarations, injunctions, orders, damages or other relief either party is entitled to.

It is first helpful to describe the programs in greater detail since this technical background is crucial to an understanding of the Research Agreement and the parties’ common background in entering into that Agreement. The starting point is the obvious need for drug companies to test their new and potentially new products for both efficiency and safety. The need arises, in part, because the drug will react with, and be processed and disposed of, by the body or with other drugs. It is therefore necessary to arrange clinical trials over an extended period across that part of the population as a whole who are potential users of the drug. Such trials are expensive and many potential drugs are discarded at some stage in the development process. It follows that any way of reliably predicting how a potential drug will react in a drug-body and drug-drug interaction within the body and across the whole of the potential population of users before and without undertaking clinical trials is of huge interest to pharmaceutical companies since this would enable many non-useable drugs to be screened out early in the development process and at relatively little cost.

The origin of the Simcyp Project was research undertaken by Professor Tucker in the mid-1980s when undertaking computer simulation studies of human drug metabolism. This led some years later to the use the techniques that had been developed to simulate metabolically-based drug-drug interactions. The first application was to simulate the interaction of Ritonvir and Methodone which are two drugs commonly used together in patients with HIV infection. This interaction has potentially harmful effects in patients taking both drugs together.

The name ‘Simcyp’ was chosen since it is a neologism derived from ‘simulation’ and ‘cytochromes P450’, the later being the major family of enzymes responsible for metabolising drugs and other foreign chemicals. The method that was developed for the Project involved obtaining in vitro biochemical data on the metabolism of drugs using subcellular factions from human live and from cell systems that had been engineered to express single human enzymes. This data was developed using the Michailis-Menten enzyme kinetic model which is able to provide predictions as to the rate of transformation of a drug by an enzyme. The resulting data was integrated with demographic data involving sex, ethnicity, weight and disease and with physiological data involving liver size and blood flow for particular human populations. The data was selected for a random population of 2000 patients by using random number generation techniques generated by recourse to the appropriately named Monte Carlo simulation methods.

In short, The Simcyp Project’s objective was to construct models which blended demographic data and physiological, ethnic and genetic features of specific patient types with in vitro data on enzyme kinetics and on the inhibiting and inducing effects of enzymes so as to provide an extrapolated prediction of in vivo, or the whole body, effects of the relevant drug-drug interactions across the range of the population and concentrations being considered. Simcyp, as developed by Professor Tucker and Dr Rostami, was a simulation approach to the prediction of specific in vivo drug-drug interactions. Although developed using a particular research-based mathematical modelling program, it was not tied to or based on any single physical software product.

The evaluation of the in vitro data had been undertaken using algorithms and databases developed by Professor Tucker and Dr Rostami. The number of equations and the volume of data that had to be developed and evaluated were such that computer assistance was required and the team turned to a particular group of source codes within MathCAD which is an advanced mathematical modelling system used for the development and testing of numerical models. It is a program which makes a good research tool since it is well suited to the rapid development of such models. The analyses provided by this system were accessible on MathCAD worksheets.

The first simulation of outcomes in patient populations using Simcyp methods was presented by Professor Tucker and his colleagues at an International Symposium on Drug Metabolism held in Grantham, England in July 1999. This presentation generated much interest amongst those who were present who were representing pharmaceutical companies. This interest was particularly generated by the realisation that the work being spearheaded by Professor Tucker on Ritonvir-Methodone interactions could provide a successful means of simulating in vivo interactions in many drug-drug pairings across the population by the extrapolation of pre-clinical in vitro studies.

MathCAD is poorly suited for routine execution by users, particularly non-academic users, who are merely interested in the computed results. Moreover, it was not readily usable by third parties and did not consist of much beyond the algorithms and equations with which it had been provided and it did not assist those simply wanting to expand its data base, its range of predicted drug-drug interactions and the population for which such predictions could be provided. For all such users, Simcyp needed a program which, unlike MathCAD, was not written in a language which was tightly coupled to a particular environment.

Dr Edwards’ task was to provide the means of transforming the Simcyp Project in the form it had reached following the predicted Ritonvir-Methodone interactions into a widely useable and readily accessible program. For this purpose, he selected Java as the language to be used. Java is an Object Orientated language consisting of a number of separate sections or classes. This is one of Java’s strengths since it can use many of its classes in different programs with little or no adaptation. A Java source file, once compiled, defines the contents of the class files. A Java class consists of data items and methods. A method is a function, routine or procedure needed to find, provide or use the data in question.

All Java programs are divided into discrete packages, each of which is built up of a number of classes. Simcyp as written in Java contained approximately 170 classes. The Simcyp packages consisted of the following:

The most complex part of the Simcyp Java Program was the GUI package comprising four sub-packages and 46 classes. Amongst these are the main application class which defines the method; the Swing package enabling the same data to be viewed differently; the Chart classes enabling charts or graphs to be drawn; the package which generates a model human population; and classes containing visual images, tables and the display of real numbers in specified format including the number of decimal places.

The development of the Simcyp program involved, initially, the drafting of a specification whose contents had to be agreed with Sheffield. This document provided the parameters, requirements and objectives of the program. The software then had to be designed, using an UML or universal modelling language, coded in Java, deployed as a useable program and then tested, revised and redeployed in revised versions.

It follows that the key features of the Simcyp Project produced by Sheffield were the algorithms and databases written in MathCAD and those of the Simcyp program produced by Dr Edwards were the Graphical User Interface or Gui and the Java Class Libraries written in Java. There was very little overlap between the two programs.

Dr Martin Gittens, Sheffield’s expert software consultant with wide experience in Java programs and modelling techniques using Monte Carlo simulations, estimated that a relatively small but significant or possibly crucial proportion of the Simcyp software code was contained in or derived from the MathCAD software. Dr Gittens explained that the linkages that occurred by way of adaptation or derivation were only to be found in some of the equations and algorithms produced by Sheffield in the MathCAD program. None of this linkage occurred in the Gui packages which comprised over half the code of the Java Program software. Within the Java Class Libraries, many of the files were unrelated to MathCAD but some were derived and some were adapted from the original MathCAD codes.

None of the files that were derived from MathCAD contained reproductions of the underlying algorithms and equations. Approximately 10% of the bytes in the Java Program were adapted from the original MathCAD codes that provided the nucleus of the Simcyp Project software. Dr Gittens explained the connection in this exchange in cross-examination:

“Q. ... the similarities between the MathCAD programs and the Java classes which you have identified in your report are simply this, that both sets of programs carry out the same task, or at least have programs in them to carry out the same task, by reference to the same algorithms and equations?

A. Yes, that has to be true.

Q. So it is the function of the software which is common to both?

A. Well, they contain the same algorithms.

Q. Yes?

A. So, yes.

Q. They are there to perform equations and the same equations are there in both?

A. Yes.”

Dr Gittens’ written and oral evidence was accepted and not challenged by Cyprotex so that this analysis was common ground between the parties. It follows that a small percentage of the files that are not associated with the Gui packages are wholly or partly adapted from Simcyp MathCAD codes in the sense that they share a common function of performing equations and contain the same algorithms. Otherwise, there is no direct connection between the MathCAD and Java programs.

It is important to keep in mind when considering the factual background to the dispute that the name “Simcyp” was used to mean three different things, being the three successive stages of the Simcyp development cycle. Firstly, “Simcyp” was used to describe the Project up to 1999 that had resulted in the MathCAD worksheets and accompanying algorithms and databases developed by Professor Tucker and Dr Rostami. Secondly “Simcyp” was used to describe the research programme in which Dr Edwards was using the Java programming language to develop the Project into a user-friendly form. Thirdly, “Simcyp” was used to describe the product in its final state at the conclusion of phase 2 of Sheffield’s development strategy when it would be ready for commercial application.

The breakdown in relations between Sheffield and Cyprotex can partly be attributed to the many occasions when the parties were at cross-purposes in considering the word “Simcyp” with the Sheffield personnel thinking of the MathCAD-based software and the Cyprotex personnel the finished Java software in a state ready for marketing. Further confusion arose since the Research Agreement only used the word “Simcyp” in its third sense as referring to the programme of Java-based research being carried out in phases 1 and 2 of its development.

The International symposium held in Grantham in July 1999 was a watershed in the development of the Simcyp Project. In the following months a number of pharmaceutical companies approached Professor Tucker and Dr Rostami and offered finance to enable the Project to continue and develop. Meanwhile, the two pioneers wrote up the Grantham presentation in the British Journal of Clinical Pharmacology that was published in December 1999 and two further presentations were made by Dr Rostami and his researchers at a meeting of the British Pharmacological Society held in Nottingham and at a seminar at Oxford University. This interest prompted the participants in the Simcyp Project to formulate a development strategy that would involve a two-phased approach. The first phase would involve the conversion of the MathCAD-based model into a user-friendly model based on a widely accessible operating system which would be readily accessible to commercial non-academic users. This phase would need, Professor Tucker and Dr Rostami estimated, up to twelve months of development work and would enable the necessary interest in the project from the pharmaceutical industry to be obtained. The second phase, with an estimated timespan of up to six months, would involve the expansion of Simcyp by the build-up of its database of information. In other words, Simcyp would be developed by first converting it from a research project into a commercially exploitable product and, once that was achieved, Simcyp would be built up into a product which was ready to be marketed. At that point, at the end of the second phase, Sheffield would be able, itself or by collaborating with or selling the product to a marketing organisation, to reap the commercial benefits of this pioneering drug development tool.

It was at this point in the development history that Sheffield and Cyprotex’s paths first crossed. The initial contact arose as a result of Cyprotex’s then work in a project which had some similarities with the Simcyp Project which had been christened ‘The Virtual Human’ Project. Cyprotex at that time was still a division of Medeval Ltd and in April 1999, a few months before Sheffield’s Grantham presentation, Dr Leahy had joined the Cyprotex Division following a direct approach from Medeval Ltd’s Chief Executive Officer, Dr Steve Toon. This approach had been stimulated by Medeval Ltd’s interest in Dr Leahy’s Virtual Human proposal which he had been working on for many years at another pharmaceutical company.

The Virtual Human aimed to provide simulated test predictions for possible new drugs as a means of saving both design time and development costs using a suite of mathematical models of the human body. The similarities between the objectives of the Virtual Human and Simcyp are obvious. Indeed, both Dr Leahy, in describing the Virtual Human, and Professor Tucker, in describing Simcyp, in their respective witness statements prepared without knowledge of the contents of the other’s witness statement, used a virtually identically worded analogy to explain the utility of their respective ‘babies’. Dr Leahy stated:

“In the same way as the aircraft industry carries out extensive simulations of how a new plane would perform, the Virtual Human would be an immensely valuable tool for improving the design of safe and effective drugs at lower cost than is currently the case.”

Professor Tucker stated, showing that Simcyp was not unique:

“Whereas the aircraft industry has used computer simulation to evaluate the performance of its products (aircraft) before they are built (and fly) for many years, the pharmaceutical industry has been slow to apply a similar principle to the evaluation of its products (drugs) before they are tested in humans for the first time. Simcyp is one of the first systematic attempts to do this: and does it uniquely with respect to incorporating information on in vitro human drug metabolism.” (emphasis added)

Professor Tucker was now very keen to develop Simcyp and market it both because of its perceived advantages as a tool for use in pharmaceutical research and for its commercial benefits to Sheffield. He and his colleagues had spent over ten years developing the means of drug-drug interactive in vitro predictions and had developed a MathCAD-based program as a viable launching pad for such a project. However, it was clear that his Department had insufficient resources or funds to take the project forward on its own since Sheffield would need to employ a full time skilled programmer to undertake the necessary programming work to develop a user-friendly program.

Sheffield had a university-based Department, its Research Consultancy Unit, which later became the Technology Transfer Office (“TTO”), which negotiated outside contracts, sponsorship and funding for the whole of the University and it costed out the proposed project at a much higher sum than Professor Tucker would be able to obtain sponsorship for and added a second blow to the effect that the Project would have to reimburse Sheffield for the time and cost of the programmer it had been planned that Sheffield would be providing.

Whereas Professor Tucker and Dr Rostami had calculated that the cost of employing a programmer would involve payment of a salary of about £34,000, the TTO indicated that, by applying Sheffield costing guidelines to the project, a sum of £130,000 was required to employ a programmer with post-doctorate experience for eighteen months. This larger figure was more in tune with Cyprotex’s subsequent estimate of up to £170,000 as the cost of employing an appropriately qualified programmer for eighteen months.

It became clear to Professor Tucker that the only way to proceed was to attract financial sponsorship from outside bodies who would be interested in this project, particularly pharmaceutical companies involved in drug research and technically-based research and consultancy organisations in the drug field. Thus, the development exercise would be paid for by sponsors attracted from the pharmaceutical industry who would form a consortium which would promote and sponsor the necessary outstanding research and development.

It was clear to Professor Tucker and Dr Rostami, therefore, that Sheffield needed to raise at least about £70,000 in cash from sponsors and, in addition, to find a sponsor who could provide a programmer or the salary instead, at an estimated additional cost of £34,000.

Dr Toon had known Professor Tucker quite well for many years and on 10 May 1999, at about the time that Dr Leahy joined Medeval Ltd, Professor Tucker had accepted an invitation from Dr Toon to join the Scientific Advisory Board of Cyprotex. It followed that Dr Leahy’s joining Cyprotex and its’ early work on the Virtual Human, Sheffield’s Grantham presentation, Professor Tucker’s arrival on Cyprotex’s Scientific Advisory Board and the realisation by Professor Tucker that Sheffield’s financial shortfall dictated the need to persuade an outside organisation or company to provide an experienced programmer at little cost to Sheffield to undertake the Simcyp programming all occurred at about the same time. It was therefore no coincidence that in about September 1999, Dr Leahy and Dr Toon went to see Professor Tucker and Dr Rostami and Dr Rostami gave Dr Leahy a full account of his work to date on the Simcyp project. It must have been an exciting and stimulating experience for both of them.

Dr Rostami explained to Dr Leahy how Sheffield intended to fund the future development work that would be needed to convert the MathCAD-based Project into a user-friendly form. Fortunately, Dr Leahy was greatly excited by the Simcyp Project. He could see that both the Virtual Human and Simcyp Projects used a similar approach, similar organisms, organs, enzymes and drugs and similar interactions governed by the same Java Classes. As Dr Leahy saw the possibilities, Simcyp could provide programming skills acquired from the Virtual Human Project to Sheffield in return for some financial return. The benefit to Sheffield would be the informed programming assistance from an organisation who had already mastered some of the relevant Java-based problems in working in a similar field. The benefit to Cyprotex would be the use of Simcyp within the Virtual Human project and the possibility of being able to use the Simcyp database and acquiring marketing rights in Simcyp.

As a result, Cyprotex was available at just the right time to offer much-needed programming services to the Simcyp Project. This offer followed a full presentation of Simcyp to Cyprotex by Dr Rostami under the terms of a confidentiality agreement and a meeting at Sheffield on 14 December 1999 at which Dr Leahy and a colleague met Professor Tucker and Dr Rostami. At that meeting, Dr Leahy persuaded the two Simcyp pioneers that it was both necessary and desirable to rewrite the MathCAD program in Java. One reason advanced by Dr Leahy and accepted by Professor Tucker and Dr Rostami was that such a Java-based program would naturally complement and be compatible with Cyprotex’s Virtual Human program.

Soon afterwards, on 19 December 1999, Dr Leahy submitted a proposal to Professor Tucker and Dr Rostami which he had prepared. The proposal was, in summary, that Cyprotex programmers would carry out the software development of Simcyp under the direction of Cyprotex and would implement phases 1 and 2 of Sheffield’s proposed development of the Simcyp Project. Cyprotex would pay the salary of the additional member of staff that would be required but would share the risks of the development in return for negotiating rights to seek to acquire exclusive rights to the commercialisation of Simcyp once the initial research collaboration with the sponsors had been satisfactorily completed. Sheffield would contribute £25,000 towards Cyprotex’s estimated cost of the programmer of £41,150. One particular reason for Cyprotex’s offer being expressed in this way was that the Cyprotex Division had no available funds to provide sponsorship money but Dr Leahy, particularly since the management buyout of the Division was imminent, was keen to participate in a project that was particularly complimentary to the Virtual Human Project.

It followed that the proposal invited Sheffield to contribute £25,000 towards the cost of employing a programmer for the first 12 months of that programmer’s work. Both Dr Leahy and Professor Tucker accepted in their evidence that at some stage they agreed that Cyprotex would be paid £12,500 for providing programming services. That agreement must have been reached informally between them soon after the proposal was sent out by Dr Leahy. The two must have discussed the proposal and agreed to split Sheffield’s proposed fee in half.

This proposal was discussed by Dr Leahy with Dr Roberts of the TTO in a telephone conversation on 22 December 1999. Dr Roberts wrote a note on her hard copy of an e-mail she had sent to Dr Leahy on 17 December 1999 which read:

“JR spoke with David Leahy. Due to difficulty of employing a suitable candidate they would still prefer to be the employer. They have no problem with IP going to Uni. Simply see the project as conversion of IP software to a more marketable format. See the involvement of other parties as feedback from customers to develop the programme for their needs. Support they will provide is in kind support. They want first option for an exclusive license to market - with a royalty return.”

Dr Leahy insisted in cross examination that when he was referring, in this conversation, to “the IP going to” Sheffield, he was referring to the algorithms and data bases making up the current MathCAD program. Moreover, in his mind, owning the IP in the Java program was different from owning the copyright in that program. He thought that it was insufficient for Cyprotex to own the IP in Simcyp to enable Cyprotex to market the Java Simcyp program once it had been developed. For marketability, Cyprotex needed the copyright, which it held, whereas Sheffield merely retain the intellectual property in that program.

I cannot accept Dr Leahy’s explanation for his comments to Dr Roberts in this conversation. The ownership in all underlying rights in the MathCAD program was never doubted or disputed and it would have made no sense for him to have agreed to such rights “going to” Sheffield. He was referring to future rights of some kind “going to” Sheffield. These rights were obviously rights in the future program they were discussing. In context, there was no difference between IP rights and copyright rights. Whatever was meant by the IP rights that would go to Sheffield, such rights clearly encompassed copyright. That is how Dr Roberts understood Dr Leahy and, as she stated in cross examination, it was what she intended. The words that Dr Roberts attributed to Dr Leahy in her note were clearly accurately paraphrased in that note and, objectively, he must be taken to have stated what the note records of him. That record is to the effect that Cyprotex would agree to Sheffield having or retaining the IP rights, to include copyright, in the Simcyp program once developed by the programmer Cyprotex hoped to employ.

Not long after Sheffield received this proposal, Professor Tucker accepted Cyprotex’s offer to become chairman of its Scientific Advisory Board that he had joined 8 months earlier but which had yet to meet. His first meeting both as Chairman and as a member of the Board took place on 12 January 2000 and it turned out to be the only meeting he attended. However, Dr Leahy took the opportunity of Professor Tucker’s presence at Cyprotex’s premises to explain in detail to him how the Virtual Human would be built and how some of the proposed Java Classes for use in Simcyp could be used in both projects. He also explained that any Virtual Human programming that would be undertaken would, whilst phase 1 of the Simcyp work was being carried out, be undertaken in parallel with the Simcyp programming.

The shape of the Simcyp development soon emerged. Sheffield produced the first draft of a Research Agreement on 22 December 1999. This draft set out the substance of the proposed method of working which never changed. The Research Agreement provided for several sponsors to make available a lump sum, the sums then being discussed varied between £20,000 and £30,000. These sponsors would all be joint parties to the Research Agreement with Sheffield. Cyprotex would also be a sponsor and would provide a programmer to undertake the necessary programming work to convert the MathCAD programme into a user-friendly Windows-based programme. The cost of this programmer would be provided partly by Cyprotex, in lieu of providing a sponsor’s cash payment to Sheffield, and partly by Sheffield who would pay Cyprotex an agreed sum of £12,500 out of sponsorship money raised from the sponsors other than Cyprotex. Both Professor Tucker and Dr Leahy understood each other’s financial constraints, in Sheffield’s case that the project could only proceed if about £70,000 in sponsorship money from outside sources plus the services of a programmer at no cost to Sheffield were available and, in Cyprotex’s case, that no cash could be made available since Dr Leahy had none at his disposal.

Although Dr Leahy and Professor Tucker understood that the programme to be written by Cyprotex would be written using Java as the programming language, that was not a pre-requisite of the proposal contained in the draft Research Agreement. It was also understood between these two that the programmer to be supplied by Cyprotex would be the same programmer as would be working on the Virtual Human. The resource of a programmer provided or funded by Cyprotex would constitute its contribution to, and its sponsorship of, the Simcyp Project.

Sheffield then spent several months attempting to obtain sufficient sponsorship to enable the Project to proceed. Dr Leahy was keen to proceed and to discuss the commercialisation of Simcyp that would, he hoped, be undertaken by Cyprotex once the software had been developed but although Professor Tucker expressed interest in Cyprotex having the commercial rights in Simcyp, he was not prepared to discuss what would happen after Simcyp had been developed when he was still attempting to finalise the necessary sponsorship for the development phases of the Project.

Thus, little of consequence occurred until June 2000. By then, Dr Leahy had discovered Dr Edwards who had worked for some time as a pharmacokineticist and he was introduced to Dr Rostami on 9 June 2000. Dr Leahy suggested to Dr Rostami that Dr Edwards should start work immediately since time had slipped, even though the Research Agreement had yet to be signed. This lead-in time would allow Dr Edwards to obtain an understanding of the MathCAD program and to help Sheffield to define its requirements for the functionality of the interface. Dr Rostami readily agreed to this suggestion since he was already aware of Dr Edwards’ reputation and was most impressed with his ready ability to grasp the essentials of the MathCAD program at their first meeting.

There was little or no discussion about the basis upon which Dr Edwards would be starting this work pending finalisation and implementation of the Research Agreement. It appears however to have been assumed by Dr Rostami and Professor Tucker on Sheffield’s side and by Dr Leahy and Dr Edwards on Cyprotex’s side that any work undertaken by Cyprotex would be undertaken as part of Cyprotex’s obligation to sponsor the Simcyp Project through the Research Agreement.

The only other clear evidence of the parties’ intentions as to the basis upon which Simcyp programming work would be undertaken by Cyprotex, other than from the wording of the Research Agreement itself, is provided by an e-mail sent by Dr Leahy to Dr Roberts of the TTO who was responsible for drawing up and agreeing the terms of the Research Agreement. This e-mail was sent on 7 April 2000 and stated that Cyprotex would hire the necessary programmer as soon as the Research Agreement was agreed and would invoice Sheffield for the first six months salary to a limit of £12,500. This understanding survived the second draft of the Research Agreement sent out by Dr Roberts on 8 May 2000 since it is set out in the Appendix which defined Cyprotex’s sponsorship terms. However, following the meeting at which Dr Edwards was introduced to Sheffield, the parties had clearly revised the wording defining Cyprotex’s proposed sponsorship contribution since the subsequent fourth draft provided that Cyprotex would provide a contribution of £17,000 and the Research Agreement itself, prepared soon afterwards, also provided for Cyprotex’s sponsorship contribution as being £17,000 without defining further how this contribution was to be provided. Cyprotex’s contribution was fixed by reference to half the cost of the salary of a programmer with the requisite experience as calculated by Professor Tucker. The other four sponsors each sponsored £15,000 so as to provide a total sponsorship sum of £77,000.

This change in wording in the draft Research Agreement arose in the light of both Dr Leahy’s and Professor Tucker’s view that each clearly held at the time Dr Edwards started work in June 2000 with the Research Agreement still unsigned, namely that Cyprotex was providing sponsorship by the provision of a programmer out of its own resources but with a cash contribution from Sheffield of £12,500 to be paid out of sponsorship money paid by other sponsors. For the purposes of the Research Agreement this provision of sponsorship in kind was reflected by providing for a monetary figure of £17,000 and by the additional understanding reached several months previously that Sheffield would contribute a £12,500 payment to Cyprotex against its programming costs out of sponsorship money once this had been received by Sheffield.

An invoice was in fact sent by Medeval Ltd to Sheffield dated 20 September 2000 in the sum of £12,000 (not £12,500) for: “software development as contracted” at about the time that the final version of the final draft of the Research Agreement was sent out by Dr Roberts to the individual sponsors. This sum plus VAT was paid on 22 January 2001 by Sheffield to Medeval Ltd.

The Research Agreement did not become operative until 21 December 2000 when the last of the sponsors’ signatures was added to the document. The version of the Agreement which was signed was the fifth draft which had been put into circulation when it was sent by Sheffield to Dr Leahy on 7 September 2000. In the meantime, Dr Edwards had started work on phase 1 of the programming work. He started work on 12 June 2000 immediately after being handed discs containing the MathCAD program by Dr Rostami, initially familiarising himself with the program and in early drafting work on the specification for the Java program he was to write. He worked on this programming, largely from his home in Loughborough, Leicestershire.

The company that became Cyprotex Ltd was incorporated on 21 March 2001 as Inhoco 2306 Ltd and that company changed its name to Cyprotex Ltd on 5 June 2001 following a management buyout of the Cyprotex Division from Medeval Ltd led by Dr Leahy. The Research Agreement was actually signed by Cyprotex over 6 months before this incorporation in the name of Cyprotex Ltd by Dr Leahy. The parties did not turn their minds to the formalities arising from Dr Leahy signing the Research Agreement in the name of Cyprotex Ltd at a time when no such company had been incorporated. Thus, Dr Leahy was either signing the Research Agreement on behalf of Medeval Ltd, or in his own name or on behalf of a company which had yet to be incorporated.

Sheffield was aware from an early stage that Dr Leahy was representing an entity then known as Cyprotex Ltd but which was in reality a Division of Medeval Ltd and that that entity was soon to be the subject of a management buyout led by Dr Leahy which would then trade as a separate company. Indeed, Dr Roberts stated in her witness statement that:

“I recall that David Leahy made it clear that he was representing Cyprotex and not Medeval and that Cyprotex was a separate operating division from Medeval. David Leahy suggested that Cyprotex were about to split away from Medeval and was looking for outright investors to assist in this process. ... Therefore I made no reference to Medeval in any of the draft Research Agreements that I sent to David Leahy and I was not corrected on this point with any of the subsequently amended agreements which I later sent to Cyprotex.”

By an agreement dated 1 June 2001 Medeval Ltd agreed to sell to Inhoco 2306 Ltd the business and assets of the Cyprotex Division. This agreement was made 4 days before Inhoco 2306 Ltd changed its name to Cyprotex Ltd on 5 June 2001. This agreement, in clause 2.1(a)(i), assigned to Cyprotex Ltd all intellectual property rights owned by Medeval Ltd relating to the business of the Cyprotex Division.

Dr Edwards explained in evidence what was involved in developing software. Each of the two phases that was envisaged for Simcyp would involve five logical and progressive steps being: the analysis of Sheffield’s requirements and the drafting of the necessary specification; the design of the software; the coding of the software in Java; the deployment of the software as a useable program; and the testing, revising and redeployment of the software as a useable program.

Dr Edwards immediately started work on the first and second progressive steps for phase 1. He started by designing and preparing prototype code segments whilst simultaneously drafting a detailed software requirements specification which was approved by Dr Rostami and Dr Leahy on 16 August 2000. In order to draft a specification, it is customary for the programmer to act as an interrogator of the customer in order to extract the precise requirements of the customer. Both parties take an active role in analysing the customer’s requirements and the customer will attempt to formulate and reformulate its requirements from the often nebulous concepts that emerge from the programmer’s specifications. In turn, the programmer acts as both an interrogator and a problem solver. This two-way exchange occupied an extended period of time during which Dr Edwards posed lists of prepared questions to Dr Rostami and Professor Tucker and listened to the needs of Sheffield that were explained to him in the words of both pharmacokineticists. Dr Edwards also analysed written and mathematical material to extract words, phrases, key objects, processes and data.

Dr Edwards undertook his analysis using material obtained from Sheffield including two MathCAD worksheets. However, these worksheets were not simply transcribed into the program being developed. The two that had been provided were for two specific pairs of drugs: Ritonvir with Methadone and Ritonvir with Viagra in AIDS patients. In consequence the worksheets did not identify a unique algorithm applicable to all interacting drugs but merely two different applications with particular assumptions and approximations. The two worksheets also showed differences of approach to in vivo scaling factors. They were solely concerned with simulations for AIDS patients whereas Simcyp was initially developed for healthy patients. There were other differences identified by Dr Edwards between the worksheets and the technique being developed in the Simcyp program for in vitro drug-drug simulations. In using the worksheets and in analysing the differences between them and the proposed program, Dr Edwards developed models which anlaysed readily available literature and interpretations and assumptions made by him that were drawn from his review of the material and information provided to him which were set out in tables of values and other documents he drafted for Sheffield’s consideration,

The specification itself is a detailed document which describes the underlying scientific, in vitro, population, trial and use case models being employed and then defines the functional requirements in terms of user and library inputs, outputs, data transformations and user interfaces and the nonfunctional requirements in terms of design constraints and documentation. It finally provides a list of definitions, acronyms and abbreviations.

The software design and coding steps undertaken following the agreement to the specification were all undertaken by Dr Edwards without either supervision or input from Sheffield. The tasks were extremely complex and involved the preparation of UML diagrams in order to provide pictorial representations of the logical steps to be followed by the coding exercises. The coding involved the development of reusable code units or classes into packages. Some of these classes were in the form of class libraries which are grouped classes or packages which are capable of use for other applications. Ultimately, Dr Edwards developed ten packages, one of which was a class library licensed from a company called Sitraka Inc. and the others being packages developed by Dr Edwards himself. Within each package was a wealth of complexity which was illustrated schematically by a mosaic of class diagrams which Dr Edwards also prepared.

This design and coding process was undertaken in two phases, corresponding to the two phases of development envisaged by Sheffield initially and by the definition of the research programme contained in the Research Agreement. He produced the first prototype product, Simcyp P1.1.1. followed by an initial revision on 16 January 2001 and further minor revisions on 31 January and 13 February 2001. The first of these minor revisions included a patch for a software bug he had found in one of his algorithms during his own testing procedures. This marked the end of phase 1. There was then a meeting of the Simcyp Sponsors on 29 March 2001 during which new features for the next version of the software were suggested of which four were chosen. This necessitated much work to prepare a new Software Requirements Specification which was dated 1 August 2001 and a new product release, version Simcyp P2.1.1. The Specification was prepared in two sections with the input data in a separate document to allow Dr Rostami more time to produce the appropriate values.

In addition to the design and coding activities, Dr Edwards undertook the necessary deployment activities for a new program including the bundling of the executable program’s classes into jar archives, the obfuscating of the executable code to hinder unauthorised reverse engineering, the development of an appropriate setup program to install the software in a user-friendly fashion, the bundling of release notes and validation files and packaging the resulting files into a form appropriate for e-mail transfer, web downloading or compact disc supply. Dr Edwards also built into the program some features to facilitate the complex simulation testing that would be required. He also ran through a series of test plans.

The work undertaken by Dr Edwards of development and coding was done largely from home and entirely on his own. Dr Rostami acknowledged this in this exchange in cross-examination:

When the relationship between Sheffield and Cyprotex broke down and Dr Edwards suspended work on phase 2 of the program, he registered Simcyp P2.1.1. as a documented milestone. This involved archiving an identifiable snapshot of the source code at that time, updating the UPL design diagrams and release notes to refer to the then current state of the software and completing preliminary tests so that they related to an identifiable version of the source codes.

The breakdown in relations between Sheffield and Cyprotex slowly developed as a result of on-going discussions between Dr Leahy and both Professor Tucker and Dr Rostami as to who owned the copyright in the phase 2 Java Classes, the related question of the wording of the license under which Cyprotex could use the Simcyp software and as to the development rights in phase 2 and the commercialisation rights in the finished product. This disagreement grew during the lengthy but leisurely exchanges between Dr Leahy, Dr Roberts in the TTO and Professor Tucker from the time when the Research Agreement was finally implemented in January 2001 until the final rift in early September 2001. Agreement was not reached on any of these topics. Relations between Sheffield and Cyprotex finally broke down, largely as a result of Cyprotex’s realisation that Sheffield planed to attempt to market Simcyp itself through a new company it had formed without Cyprotex’s involvement and would be claiming intellectual property rights in the MathCAD algorithms and data sheets and, it was feared by Dr Leahy, in the Java Classes as well.

The final breakdown occurred when Dr Edwards informed Dr Rostami that he had been instructed by Dr Leahy to suspend support for the Simcyp development. By then, as Dr Edwards saw the situation, he had been held up by a lack of necessary data from Dr Rostami. This e-mail was sent on 12 September 2001. Cyprotex started these proceedings soon afterwards on 27 September 2001.

Events moved on rapidly following the breakdown of commercial relations between Sheffield and Cyprotex. None of the Sponsors received any further versions of Simcyp and Sheffield did not return any of the sponsorship money to any Sponsor. Sheffield, according to Professor Tucker, has since reprogrammed the whole system. However, Sheffield still wish to assert its perceived entitlement to the Simcyp copyright since that would give Sheffield more freedom in the future to develop add-ons to the Simcyp program. However, since the new program has been written in C ++, the Java classes would not be of much use for bolting on to any such further developed Simcyp program.

Cyprotex wish to assert its perceived entitlement to use the Simcyp copyright because the program consists of re-usable modules or building blocks which would be useful in Cyprotex’s Virtual Human project and in other possible projects as well. This potential utility arises because Cyprotex’s on-going work in these fields will include simulations as to how a drug distributes into different parts of the body and further work investigating how to set up a clinical trial involving simulations and the use of in vitro data. This work would not involve the use of Sheffield’s algorithms or data bases nor the MathCAD program provided initially to Dr Edwards and from which his work on the Java program development and code writing had been developed.

It is necessary to set out much of the Research Agreement:

“RESEARCH AGREEMENT between the University of Sheffield, and ‘The Sponsors’ ... . WHEREAS, the research programme contemplated by this agreement is of mutual interest and benefit to the University and to the Sponsors

NOW THEREFORE, the parties agree as follows:

1.

SPONSORS. As listed in Appendix 2.

2.

STATEMENT OF WORK. The University shall perform the “Programme of Research” entitled “SIMCYP - a Windows Based Simulation Program to Assess the Likelihood of Metabolic Drug-Drug Interactions from In-Vitro Data as described in Appendix 1. The University agrees to liaison meetings with the Sponsors as mutually acceptable to provide project progress information.

3.

PRINCIPAL INVESTIGATORS. The research will be conducted by Dr Amin Rostami, Professor Geoff Tucker and Dr Martin Lennard.

4.

PERIOD OF PERFORMANCE. Notwithstanding the date of this agreement, the “Programme of Research” shall be conducted during the continuous period of 12 months starting not later than 1 March 2000 unless otherwise agreed and will be subject to renewal only by mutual agreement of the parties.

5.

PAYMENT AND PRICE. The Sponsors will pay the University the sums as set out against the relevant Sponsor’s name in Appendices 3, 4, 5, 6 and 7.

6.

TERMINATION. Each Sponsor may withdraw from this agreement upon sixty days written notice given to the University and the other Sponsors. This agreement may only be terminated by the University upon reasonable notice and as soon as practicable as soon as it has been determined that circumstances beyond its control make continuation of the Programme of Research impossible on the basis contemplated by this agreement. In the event that all Sponsors withdraw from this agreement and it is terminated, the University will be reimbursed for all reasonable costs contemplated by this agreement ... and each Sponsor shall only be responsible for a proportion of such approved costs commensurate with the proportion of that Sponsors’s payment of the total project cost as set out in the financial appendices and in any event, no Sponsor shall be liable for such incurred costs in excess of that Sponsor’s agreed payment as set out in the financial appendices nor shall the University be entitled to any payment in excess of the total payment in excess of the total project price specified in Article 5 and the financial appendices. …

7.

LIABILITY.

(a)

The University will exercise reasonable skill and care to ensure the accuracy of the advice, information and drawings provided in connection with the Programme of Research but the University will not accept any liability whatsoever in respect of any claim or claims arising from the use by the Sponsors or by any third party of any such advice, information or drawings.

(b)

The University shall use its best endeavours to ensure that it will not infringe any third party rights in performance of the Programme of Research and the rights granted to the Sponsors herein. ... the University does not accept any responsibility whatsoever for infringement of such rights.

8.

SCIENTIFIC PUBLICATION The Sponsors recognise that the University will wish to publish certain papers, articles, books, pamphlets or published material which makes reference to or contains material arising from the Research Programme and it is the Sponsors intent not to hinder the normal exchange of scientific information which occurs in journals or at scientific meetings. The University agrees that the written permission of each Sponsor will be obtained before such publication or disclosure and the Sponsors will respond to such written requests within 60 (sixty) days of receipt thereof. Such permission will not be withheld longer than is necessary to enable arising intellectual property to be protected and in any event not longer than six months from the date of the University’s request for permission to disclose. Title to and the right to determine the disposition of any copyrights or copyrighted written material first produced or composed in the performance of this research shall remain with the University, provided that the University hereby grants to the Sponsors an irrevocable, royalty-free, paid up, non-exclusive right and licence to reproduce, translate and use all copyright material for its own purposes. The Sponsor has the right to assign such aforesaid rights and licence to its affiliates.

9.

INTELLECTUAL PROPERTY RIGHTS

(a)

All intellectual property in any form owed or existing at the date of this agreement and used in connection with the Programme of Research (“Background IPR”) shall remain the property of the party introducing the same.

(b)

“Resulting Intellectual Property” shall mean individually and collectively all inventions, improvements and/or discoveries whether or not patentable or capable of other intellectual property protection which are conceived and/or made by one or more members of other agents of the University acting either on their own or jointly with one or more employees of the Sponsors in performance of the Programme of Research and relating to its objectives.

(c)

All rights to Resulting Intellectual Property under the Programme of Research shall belong in the first instance to the University.

(d)

Rights to inventions, improvements and/or discoveries, whether or not patentable or capable of other intellectual property protection, relating to the Programme of Research made solely by employees of the Sponsors shall belong to each Sponsor respectively (“Sponsor IPR”).

(e)

The University hereby grants to each Sponsor a non-exclusive, world-wide, irrevocable, royalty-free licence to use the Resulting Intellectual Property for the purposes of the Sponsors internal research and development in support of the Sponsors own business activities including, but not limited to, dealings with any regulatory authority. The Sponsor has the right to assign or sub-licence such aforesaid rights and licence to its affiliates. To the extent any Sponsor IPR is also requested by a Sponsor to obtain the full benefit of this licence, each Sponsor hereby grants a similar licence to the others as specified in this clause 9(e)

(f)

The University shall have the right to grant licences to third parties under the Resulting Intellectual Property provided by the University ensure that the rights of each Sponsor under this Agreement are fully protected and in any event, any such licence shall not affect the licence granted to Sponsors set out in clause (e) above.

(g)

In addition to the licence to be granted to the Sponsors in clause 9(e), the University shall grant to each Sponsor, to the extent that it is free to do so, a nonexclusive licence to use all the University’s Background IPR where such Background IPR is either

(i)

incorporated into the Programme of Research

or

(ii)

needs to be licensed to Sponsors in order to enable Sponsors to practice lawfully any patent, invention, design or improvement arising from the Programme of Research.

Any licence granted to the Sponsors under this sub-clause 9(g) shall be world-wide, irrevocable and royalty free for the purposes of the project and for the purposes of the licence granted under clause 9(e) and thereafter royalty bearing on reasonable terms and conditions to be agreed if the Sponsors wish to use the Resulting Intellectual Property for its own commercial purposes involving the sale or licensing of the software to be developed in the Programme of Research.

10.

USE OF NAMES AND CONFIDENTIALITY

10.1

Neither party will use the name of the other in any form of publicity without the written permission of the other.

10.2

The receiving party undertakes to treat as strictly confidential and not to divulge to any other third party, in whole or in part, any information, technical knowledge, know-how, experience, data and business background of a secret and confidential nature relating to the Programme of Research (“Information”) disclosed under this agreement and not to make use of any such Information without the disclosing party’s prior written consent. The receiving party also agrees not to reverse engineer or otherwise analyse any of the materials provided under this agreement. The receiving party further agrees not to reproduce any Information disclosed to it by the other party except for the purposes set forth in this Agreement above and shall limit disclosure of Information to those persons employed or otherwise under the contract with the receiving party who have a legitimate need to know such Information for the purpose set forth herein. Such obligations of confidentiality shall not apply in respect of information that

i)

was already or comes to be in the public domain or

ii)

is already know to the recipient or

iii)

is acquired by the recipient from third party sources or

iv)

is acquired by the recipient from its own independent research or

v)

is approved in writing for disclosure by the disclosing party or

vi)

is required to be disclosed pursuant to any legal or regulatory requirements.

…

13.

GENERAL

(a)

This agreement and the documents referred to in it form the entire agreement between the parties relating to the subject matter and supersedes all previous agreements (if any) relating to its subject matter.

(b)

A waiver by any party of any term or condition of this agreement in one instance shall not be deemed or construed to be a waiver of such term or condition for any similar instance or of any subsequent breach. All rights, remedies, undertakings and obligations herein are cumulative.

(c)

This agreement may only be amended by a further written agreement duly signed by or on behalf of each party.

(d)

Nothing in this agreement shall create or be deemed to create a partnership or relationship of principal and agent between the parties.

…

Appendix 1

PROGRAMME OF WORK

The programme proposes to create a Windows based user friendly software that uses latest information on in vitro

- in vivo extrapolation to predict metabolic drug-drug interactions.

There are a number of techniques (using software equations) which are used to do in vitro - in vivo extrapolation. Our software will include number of these options in a library of models which can be selected and employed by the end user. Sponsors, depending on the level of support, may ask for specific routines of extrapolation (common to their practice) to be included in the library.

University will advertise and recruit a suitable computer programmer with some knowledge of modelling/simulation to produce the software.

The first step in the programming will be to convert prototype of SIMCYP (written in MathCad) to Windows based program. Upon successful completion of the first stage (5-6 months) sponsors will be invited to review the programme and draft their suggestions and include additional features that they would like to incorporate into the software. These will be considered by principal investigators and efforts will be made to incorporate such additional features. However, the proportion of additional features from the list of suggested features will depend on the relative contributions received from each sponsor as well as scientific validity of such features.

A second review of the software will take place at a later stage (8-10 months) and final comments on improvements/enhancements of specific aspects will be received from the sponsors. The programme will be completed by 12 months and the software at this stage will be considered as the final product. Any additional work on software from that point will depend on reaching an agreement for extension of contract between all (/part of) sponsors and the Univ of Sheffield.

…

Appendix 5

Sponsor Cyprotex Ltd

Contribution £17,000 sterling

Payment terms Six months after signature

…”.

It can be seen that the Research Agreement appears to envisage that Sheffield would conduct a Programme of Research whose principal objective was to undertake the programming work required by the Research Agreement which would be undertaken by a specially recruited employee of Sheffield. Each sponsor would contribute a defined sum at the time stipulated in the Agreement, that each sponsors’s liability to contribute to expenses in the event of early termination and to ask for improvements and enhancements in the program would be governed by the relative contributions of each sponsor but with liability capped by the sum sponsored and that any additional work on the software would depend on an extension of the Research Agreement being agreed with Sheffield.

The intellectual property rights in the program are defined in detail in clause 9. Sheffield was to hold intellectual property rights in anything used in connection with the Programme of Research owned or in existence at the date of the agreement (which date was not defined) and any improvements and discoveries conceived by Sheffield solely or jointly with one or more employees of the sponsors but that any sponsor was to retain rights to improvements made solely by them or their employees. Thus, the agreement did not expressly deal with intellectual property rights in improvements resulting from programming work undertaken by a sponsor.

I will first summarise the disputes to which the law of copyright must be applied. The first dispute arises out of the existence of the Research Agreement which, on its face, appears to provide in some detail for copyright ownership in the Java-based program. However, as a result of the relationship between Sheffield and Cyprotex before the Research Agreement came into effect and of the wording of the Research Agreement itself, Sheffield contends that the Java programming work was not carried out under the aegis of the Research Agreement at all but under an informal prior engagement which arose at Sheffield’s behest to enable it to fulfil its obligation to perform the Programme of Research owed to all Sponsors and which was provided for in the Research Agreement.

Once the applicable contract has been identified, I must then determine whether copyright ownership is provided for in that contract or whether, in the absence of clear express terms providing for ownership, an appropriate term covering ownership should be implied or whether copyright in the Java program is to be determined by the general law of copyright.

I must also determine who is entitled to claim copyright in the Java program. This involves deciding three related but separate matters. Firstly, I must decide whether the Java program is a single and indivisible whole or is a series of discrete classes to which copyright must be allocated separately. Secondly, in relation to the whole program and for any discrete part of it, I must decide whether the mix of MathCAD and Java is such that the resulting program is subject to the pre-existing MathCAD copyright or to a separate copyright arising out of the authorship of the Java-based program. Thirdly, I must decide whether the work of Dr Rostami, in collaborating with, supervising and providing input into, the Java programming exercise of Dr Edwards, made him a joint author of the Java program with Dr Edwards.

In the light of these decisions, I must then either apply the relevant and applicable express or implied contractual terms or the general law in determining whether, and if so what, declarations as to copyright are appropriate. In doing so, I must also determine whether the question of copyright ownership of the Java Simcyp program written by Dr Edwards remains a live question or whether, as appeared to be a possible conclusion, neither party intends to make any further use of that Java Simcyp program so that the sole remaining purpose of seeking from the court a resolution of the dispute as to the copyright ownership of the Java Simcyp program is to resolve who should be responsible for the parties’ litigation costs of this dispute. If this is the case, I would have to decide whether it was appropriate for the court to grant a declaration of ownership at all.

I will first consider whether Sheffield owned or retained Copyright in the Simcyp program by the application of the general law of copyright without recourse to the Research Agreement. Copyright in artistic works, including computer programs and their associated codes and specifications, is vested in the author unless the author agrees in a written contract of a recognised type that it should vest in another. What is protected is the skill and labour involved in the artistic or technical endeavour involved in creating that work. Thus, copyright protection is concerned with the process of the creation of a work and not with the ideas going into it nor with the functionality or end-product that results. In computer programming, therefore, what is capable of protection are the codes that are written by the programmer including the design and structure of the program.

A computer program often involves, as the Simcyp software development showed, a considerable interplay between a number of very technically gifted people. However, that involvement would only give the participant an entitlement to copyright protection, whether as a sole author of the parts written by the collaborator or as a joint author in the whole of that program, if the participant has written, or contributed to a significant extent in the writing of, the program or a discrete part of the program by way of authorship or joint authorship of any or all of its constituent packages, classes and libraries.

In this case, Sheffield contends that writing the Java Simcyp program involved Dr Edwards copying and reworking the MathCAD program. Thus, it contended, Sheffield retained copyright in at least those parts of the Simcyp program written by, or derived from the work of, Dr Rostami However, it is clear from the decision of Plowman J in Warwick Film Productions Ltd v Eisinger and Others [1967] 3 All ER 367, which applied the earlier decision of the House of Lords in Ladbroke (Football) v William Hill (Football) [1964] 1 All ER 465, that the work in question should be looked at as a whole. As Lord Reid stated in the latter case:

“... the more correct approach is first to determine whether the plaintiffs’ work as a whole is ‘original’ and protected by copyright, and then to inquire whether the part taken by the defendant is substantial.” (page 469)

Given that none of the Java files were derived from the original MathCAD program, that only a relatively small number of files or bytes in the Java program were adapted from the original MathCAD, that none of those adopted files related to the Gui packages and none contained reproductions of the underlying algorithms and equations, it was not surprising that Sheffield did not seek to argue with any vigour that Dr Rostami or his employer was, under the general law, the owner of the copyright in any part of the Simcyp program.

Sheffield also contended that the Java Simcyp program was the product of the joint authorship of Dr Edwards and the Sheffield team, particularly Dr Rostami, in devising the algorithms and data bases and writing the MathCad program, all of which were provided to Dr Edwards. It is not easy to delineate the constituent parts of and workmanship involved in the development of the Java Simcyp software into those of authorship or joint authorship of the whole or any part of the program and of collaboration, insertion, copying or assistance in any authorship which still stopped short of authorship. This is because the Java Simcyp program had such a varied and technical input. The necessary process of delineation is, however, helpfully explained by Laddie J, with his unrivalled and commanding experience in this area of the law, in Fylde Microsystems Ltd v Key Radio Systems Ltd [1998] FSLR 449 at pages 456 - 457. The issue in that case was whether the work of a collaborator in the development of software with particularly technical functionality who was regularly consulted by the programmer and who provided significant assistance in its testing and development amounted to authorship or joint authorship of that program. In explaining why not, Laddie J stated:

“In relation to authorship, it seems to me that two matters have to be addressed. First it is necessary to determine whether the putative author has contributed the right kind of skill and labour. If he has then it is necessary to decide whether his contribution was big enough. The latter issue in particular is a matter of fact and degree.”

In this case, the role of Dr Rostami was similar to that of the technical representative of the manufacturer of the telecommunications equipment for which software was being written by the defendant. In the Fylde Microsytems case, the technical representative of the manufacturer was experienced in the design and operation of radios. He assisted in error fixing, in suggesting the cause of some of the faults and in reporting faults and bugs. He helped set the specification and the parameters and timings within the software, which was part of the task of setting the specification. Finally, he provided technical information about the hardware into which the software was to be fitted. In consequence, the manufacturer’s representative was not a joint author of the program in question.

Similarly, Dr Rostami provided background information about drug-drug interactions and the process of simulating test results, about the Monte Carlo methods that had been used and about the algorithms that had been devised by him. He also provided test data and the MathCAD program which is a purely research program. He assisted in the compiling of the specification by providing detailed consideration to what Sheffield wanted the Simcyp program to contain and what its parameters should be. Finally, he added a contribution in his work of checking the phase 1 program, in suggesting what should be added for phase 2 and in vetting the suggestions of the other Sponsors. However, this was the work not of an author of the software but of a client who wanted to ensure that the program when written provided the desired functionality specified in the Programme of Work. It follows that Dr Rostami was not a joint author of the Simcyp Program.

Thus, if the question of copyright of the Simcyp Program is to be determined without reference to the Research Agreement or any preceding contract between the parties, Cyprotex would be adjudged to be the owner of that copyright. However, Sheffield’s case is based on express or implied terms of whatever contractual relationship governed the writing of this software.

The first matter to determine is whether the Research Agreement governed the entirety of the parties’ relevant relationship or only that part of the relationship that occurred after the date that the Agreement had been signed and had become effective. This date occurred about six months after Dr Edwards started programming work and not long before phase 1 of the Programme of Research was completed. It is a moot point whether the Research Agreement first became effective, so far as Cyprotex’s contractual relationship with Sheffield is concerned, when the last signature was added to the Research Agreement or when the last of these two parties’ signatures was added. On either view, the Research Agreement first became effective long after Dr Edwards started programming the Simcyp program.

The answer to this question depends on whether the parties to the Research Agreement intended, once it first became effective, that it should then have retrospective effect for the entire period of the parties’ relationship including that part that had preceded its first becoming effective.

In support of its contention that the Research Agreement had retrospective effect, Cyprotex relied on the decision of Megaw J in Trollope & Colls Ltd v Atomic Power Contractors Ltd [1963] 1 WLR 333. In that case, extensive civil engineering works were carried out by sub-contractors for the main contractors constructing a nuclear power station. Work started some ten months before the contract was agreed and entered into. Megaw J found that the preceding negotiations had been conducted on the basis that the contract would govern all the work being carried out. The contract thus governed the whole of the work including that carried out in the period before the contract came into existence. The basis of this finding was that there was an implied term of the contract that the contract conditions would apply retrospectively. The judge stated, at page 339, that:

“But, so far as I am aware, there is no principle of English Law which provides that a contract cannot in any circumstances have retrospective effect, or that, if it purports to have, in fact, retrospective effect, it is in law a nullity. ... Often, as I say, the ultimate contract expressly so provides. I can see no reason why, if the parties so intend and agree, such a stipulation should be denied legal effect.”

It followed that the negotiated contract that had been entered into by the parties, who had already been undertaking a commercial relationship for some months and whose terms covered all work including work carried out prior to the contract, had retrospective effect. There was no express term to that effect but this result was one, so the judge found, which both parties had intended during their pre-contractual negotiations.

Recently, I decided Northern & Shell Plc v John Laing Construction Limited [2002] Con LJ, a case where the relevant cause of action accrued on a date that preceded the coming into effect of the contract. In reliance on the Trollope & Colls case and four other more recent authorities, I drew the following conclusions about the potential retrospective effect of a contract (see paragraph 40 of the judgment):

There are two respects in which it is contended that the Research Agreement had retrospective effect. Firstly, it is contended that the agreement as a whole governed the entirety of the programming work undertaken by Cyprotex and secondly that when, in clause 9(a), the agreement provided that the copyright in Background IPR that a party owned was that: “owned or existing at the date of this agreement”, the Research Agreement was referring and relating back to the date on which the Research Agreement first had retrospective effect rather than to the date on which it first became effective on being signed. Although the wording of the Research Agreement envisaged that that date would be on or before 1 March 2000, commercial common sense would suggest that the Research Agreement should be read as if that date read: “1 March 2000 or such later date as the Programme of Research is commenced”.

The Research Agreement appears to distinguish between the date of the Agreement and 1 March 2000 but that is because the Research Agreement was drafted on the basis that it would take effect on a date that was earlier than the commencement of the Period of Performance or Programme of Research and that that commencement was stated to be 1 March 2000. If, as occurred, the Programme of Research commenced after 1 March 2000 and also the Research Agreement took effect after that commencement, there is nothing in the Research Agreement that precludes the two dates from coinciding so that the retrospective date from which the Research Agreement took effect and the starting date after 1 March 2000 of the twelve-month Period of Performance of the Programme of Research could coincide as 12 June 2000.

It is clear from both the surrounding circumstances and the language of the Research Agreement that both Sheffield and Cyprotex envisaged that it would take effect in advance of any programming work being started. Thus, on 8 June 2000, just before Dr Edwards had started to work for Cyprotex and the day before he was first introduced to Dr Rostami and started programming work, Dr Leahy wrote to Dr Rostami to announce the forthcoming introductory meeting with the words: “assuming that the University can finalise the agreement he can start work alongside you and Geoff [Tucker]”. The letter also stated: “Strictly speaking I should not put any effort behind this [project] until we have that [agreement] but things have dragged on and if you are willing (and I assume you are only waiting for the paper work) I would like to press on with things as soon as Duncan [Edwards] is fully on board.”

The language of the Research Agreement clearly envisaged a timescale which started with the Agreement coming into effect and then continued with the commencement of the programming work. The Research Agreement provided that Sheffield “shall” in the future perform the Programme of Research which it defined as the proposed creation of a Windows-based simulation program entitled Simcyp. The Programme of Research was envisaged to be for a continuous period of twelve months starting not later than 1 March 2000 (clause 4). Thus, although this date slipped by just over three months to 12 June 2000, the period was clearly envisaged as starting after the Research Agreement had taken effect. Finally, the Programme of Work appended to the Research Agreement envisaged that Sheffield would advertise for a programmer once it had taken effect.

It follows that the parties envisaged, and the Research Agreement provided, that the Programme of Research was to be governed in its entirety by the Research Agreement and that, although it only took effect ten months after 1 March 2000 and six months after the Programme of Research was started, the Research Agreement had retrospective effect and that “the date of this agreement” was intended to be the date the Research Agreement first became effective and also the start date of the Programme of Research, which turned out to be the 12 June 2000 following Dr Edwards’ introductory meeting with Dr Rostami.

Sheffield contended that the applicable contract was the informal contract entered into by Sheffield and Cyprotex when Cyprotex agreed to supply the programming services of Dr Edwards in early June 2000 following its proposal in December 1999 that it would provide such computing services. The agreed consideration for this work was, according to Sheffield, £12,500. Cyprotex maintained that the only applicable contract was the Research Agreement. Clause 13(a) of the agreement could not be clearer in establishing that it constituted the entire agreement between the parties and superseded all previous agreements. To this, Sheffield contended that, at least for the period prior to the last signature being added to the document in January 2001, the earlier contract prevailed and, in any case, the programming work was not work “relating to [the Research Agreement’s] subject matter” and therefore it fell outside the ambit of the Research Agreement.

Since the Research Agreement had retrospective effect, it clearly governed the entirety of the programming work and took effect at the moment that that work started on 12 June 2000. Moreover, the ambit of the Research Agreement was very wide since it covered all work “relating to the agreement’s subject matter” as stated in clause 13(a). The core feature of that subject matter was the development of a software program so that any programming work undertaken by Cyprotex in relation to Simcyp was obviously work “relating to” the Research Agreement’s subject matter. Clause 13(a) also stated that the Research Agreement superceded all previous agreements between the parties.

It follows that, for better or for worse, the entirety of the programming work undertaken by Cyprotex and the entirety of the program that was the product of that work were both covered by the words of the Research Agreement.

The party with whom Sheffield contracted must next be considered since “Cyprotex Ltd” was not in existence as a separate company at the date that Dr Leahy signed the Research Agreement for and on behalf of that entity. Cyprotex was still a division of Medeval Ltd and there is no clear evidence that that division of Medeval Ltd was trading in the name of Cyprotex Ltd at any time prior to the incorporation of the separate company called Cyprotex Ltd (albeit initially with the name Inhoco 2306 Ltd). Since the enactment of section 36C of the Companies Act 1985, a person who signs a contract on behalf of a company which has yet to be incorporated cannot bind that company even from the subsequent date of incorporation but, instead, becomes personally liable on that contract. Thus, either Medeval Ltd or Dr Leahy in his personal capacity was the contracting party.

The obvious inference from the factual background as to who was the contracting party is as follows:

It follows that Cyprotex Ltd was the contracting party and is the correct party to both bring and defend this action to enforce such rights as the Sponsor who undertook the Java-based programming work is now entitled to both bring and defend.

The Research Agreement is not happily drafted in several major respects.

Firstly, it did not cater expressly for the type of sponsorship that was provided by Cyprotex in that it stated that Cyprotex would provide £17,000 six months after signing the Research Agreement whereas its sponsorship was agreed to be provided in kind.

Secondly, it was drafted on the basis that Sheffield would employ the programmer needed to undertake the Programme of Research, yet these services were in fact provided by Cyprotex as one of the Sponsors.

Thirdly, the Research Agreement was undated and provided for a start date for the Programme of Research nearly four months earlier than the actual start date.

Fourthly, it did not cater for a situation, as occurred, in which the Period of Performance and the Programme of Research were suspended indefinitely prior to the conclusion of phase 2 of the Programme of Research without the Research Agreement being terminated by recourse to the contractual termination provisions contained in clause 6 of the Research Agreement. Sheffield has not alleged that Cyprotex’s suspension of work in September 2001, which was never lifted, amounted to a repudiation of the Research Agreement by Cyprotex and has not claimed damages or any other contractual remedy for Cyprotex’s apparent failure to conclude the necessary programming it had agreed to provide.

It follows that the construction of the Research Agreement must be undertaken on an artificial basis, if the wording is construed strictly and without reference to its underlying factual basis and commercial purpose.

The difficulty of construing and giving effect to the Research Agreement is similar to the difficulty that confronted the House of Lords when construing and giving effect to the commercial agreement in ICS v West Bromwich Building Society [1998] 1 W.L.R. 896 (H.L.(E)). In that case, the wording of the agreement in issue could not be given literal effect to if the commercial purpose of the agreement was given effect to.

I have set out in an earlier decision of my own my understanding of the approach that a commercial court should adopt in the light of the ICS case when construing a commercial contract which has not been clearly or appropriately drafted: (see Union Camp Chemicals Limited v ACE Insurance SA-NV (unreported, 5 June 2001)):

“38.

The relevant principles relating to the construction of disputed contractual terms and as to the material that is admissible to assist in that exercise are well-known, being enshrined in the frequently cited cases involving the construction of commercial contracts and of terms incorporated into them from related contracts. The following is a summary of these principles.

1.

The meaning of a document (or of any other utterance) is that that would be given to it by a reasonable person having all the background knowledge which would reasonably be available to the parties in the situation in which they were at the time of the contract. That meaning is the meaning that the parties using those words against the relevant background would reasonably have been understood to have intended.¹

2.

If something appears to have gone wrong with the language and the natural and ordinary meaning of the words used flouts business commonsense, that meaning must yield to business commonsense.²

3.

In the case of doubt, wording in a contract is to be construed against a party: (a) who seeks to rely on it in order to diminish or exclude his basic obligation; and (b) who proposed it for inclusion in the contract.³

4.

When interpreting words of incorporation, a literal meaning is not usually to be given to them. Instead, the words should be adapted or interpreted so as to give effect to the context from which they have been drawn and into which they have been inserted. This process involves an intelligent and not a mechanical approach to interpretation.⁴

5.

There is no fixed rule in any category of commercial contract that terms incorporated from one contract should bear a fixed or standard meaning in the second contract.⁵

6.

In determining whether a particular obligation has been incorporated from the primary contract where the words of incorporation are general, only those terms inthe primary contract which are germane to the rights and obligations of the parties under the secondary contract and are not ancillary to them will be incorporated into the secondary contract. ⁶

The process that I have just summarised is one involving the construction, albeit the liberal construction, of the express words of the Research Agreement. Sheffield sought to go further and to imply a term into the Research Agreement concerned with the ownership of the copyright in the Simcyp Program. Ordinarily, terms will only be implied into an agreement which are necessary to fill an obvious lacuna in the agreement and to give it business efficacy. These general principles are set out in Liverpool City Council v Irwin and Another [1977] AC 239, HL which both parties relied on. Sheffield suggested that, in this case in the context of an agreement concerned with the authorship and ownership of the copyright of software with a highly technical subject-matter, it was necessary to imply a term that Sheffield owned the copyright in the Simcyp Java-based program that had been written by Dr Edwards of Cyprotex.

The circumstances in which a term as to the ownership of copyright might be implied into an agreement providing for software development involving input from different sources so as to vest copyright in someone other than the author of the program were considered by Lightman J in Ray v Classic FM plc (unreported, 18 March 1998 but noted in 148 NLJ 445). The agreement in that case provided for the authorship of documents which were to be compiled from several sources under a consultancy agreement where the intellectual property rights in those documents were claimed by both the compiler and the company for whom the compiler was undertaking the compilations. Lightman J set out guidelines indicating the circumstances in which a term would be implied so as to grant copyright in the software to the client and not to the author thereby overriding the general law that would otherwise grant copyright to the author. Sheffield contended that if clause 9 of the Research Agreement did not expressly grant it the copyright in the Java-based program, these guidelines were applicable so as to give rise to an implication that that copyright vested in Sheffield.

Each party contended that the Research Agreement, and in particular clause 9 when properly construed, vested the ownership of the copyright in the Simcyp Java-based program in that party. Cyprotex pointed to clause 9(d) as the governing provision that determined this issue in its favour whereas Sheffield pointed to clauses 9(a), 9(b) and 9(c), when taken together, as determining this issue in its favour.

Alternatively, Sheffield contended that there was a lacuna in the Agreement since it did not provide at all for the ownership of the copyright in the Java-based program since the program had been written by a Sponsor and, in consequence, a term should be implied which granted Sheffield that copyright. Cyprotex responded to this alternative argument by contending that the usual rule should prevail that the author of a work keeps the copyright in it, unless there is an agreement to the contrary.

In construing the Research Agreement, it is helpful to deal first with two general considerations.

Firstly, it is helpful to summarise the basic underlying facts common to both parties at the date that the Research Agreement was entered into in December 2000.

Both Sheffield and Cyprotex were very keen to develop Sheffield’s MathCAD-based program. This program provided for simulated in vitro drug-drug interaction predictions using algorithms and data bases developed by Sheffield. Cyprotex wished to become involved in the Simcyp Program for two inter-related reasons: to assist it in its development of its Virtual Human project and so as to provide a means of securing valuable commercial rights from Sheffield to market Simcyp once it had been fully developed. Sheffield wished to develop the Simcyp Project for two reasons: the Department of Clinical Pharmacology wanted to take forward its pet ten-year phamacokinetics project which had already earned international renown and to obtain valuable revenue from a potentially successful commercial exploitation of that Simcyp Project. The problem for both parties was that each lacked necessary readily-available finance. Sheffield could only proceed with outside financial sponsorship and Cyprotex had no readily available finance at all. However, both of these parties accepted that if other sponsors could be found, Cyprotex could sponsor the project by providing the necessary programmer and programming skills. This resource was particularly attractive to Sheffield given Cyprotex’s obvious experience and expertise in the particular specialised programming that would be required.

The intellectual property rights in the program would be acquired by Sheffield save for those in any improvement in the Java-based Simcyp program that might be provided solely by Sponsors. The aim of the Simcyp Project was to produce a marketable program which Sheffield would either market itself or would agree commercial terms with a Sponsor or with a third party to enable that other party to market the program. The Research Agreement did not give Cyprotex any direct entitlement to those marketing rights but, instead, merely provided Cyprotex with a non-contractual expectation that it would be in a favourable position to negotiate marketing rights from Sheffield once the Programme of Research had been successfully completed. This expectation was one that the Research Agreement gave all the other Sponsors as well.

Secondly, it is helpful to bear in mind that the relevant principles of the general law of copyright governing the authorship of computer programs and software were known to and were in the minds of both Dr Leahy of Cyprotex and Dr Roberts of the TTO.

Thus, it would have been known to both parties that the author of the Simcyp Java-based program would acquire copyright in that program unless the Research Agreement provided otherwise, that joint authorship in that program would only arise if the additional input from Sheffield was substantial and consisted of authorship of the computer program and that Sheffield wanted to retain its copyright in the MathCAD program and to obtain copyright in whatever Java-based Simcyp program was produced.

With this background in mind, it is clear that the Research Agreement should be construed as follows:

Once these considerations have been taken into account, it can be seen that the Java program was intended to be covered by the definition of Resulting Intellectual Property since it fell within the definition of such property, namely “inventions, improvements and/or discoveries which are conceived and/or made in performance of the Programme of Research” provided for in clause 9(b) of the Research Agreement. Essentially, the Java-based program could be seen to be an improvement of the Background IPR but it might also be said to have contained elements which were discoveries. This conclusion arises because the prime object of the Research Agreement was to improve the Background IPR so that what resulted was improved IPR which would be commercially marketable. In other words, “Resulting Intellectual Property” was to include the windows-based software that was to be produced by the Programme of Research.

Clause 9(c) provided that all rights to Resulting Intellectual Property under the Programme of Research should belong in the first instance to Sheffield. The question for determination is whether that initial presumption was displaced by clause 9(d), which provided that rights to Resulting Intellectual Property made solely by employees of the Sponsors should belong to “each Sponsor respectively”, or was displaced because Clause 9(b) had such a limited ambit that the Java-based program fell outside the definition of “Resulting Intellectual Property” altogether.

Taken literally, the words of clause 9(d) are clearly apposite to cover Dr Edwards’ programming work, to define that work as Sponsor IPR and to vest intellectual property and other rights to the Java-based Simcyp program in Cyprotex. That clause reads:

“Rights to improvements and/or discoveries whether or not patentable or capable of other intellectual property protection, relating to the Programme of Research made solely by employees of the Sponsors shall belong to each Sponsor respectively (“Sponsor IPR”).

Against the background already set out, however, these words are inapposite to transfer copyright in the entirety of the Simcyp Java-based program to Cyprotex for four related reasons:

On analysis, therefore, clause 9(d) does not appear to cover Cyprotex’s programming work at all. This conclusion is supported by a consideration of the commercially unreal consequences of a conclusion that clause 9(d) covered Cyprotex’s Java-based programming work. These consequences would include the following:

Overall, the commercialisation of Simcyp would be virtually impossible since neither Sheffield nor Cyprotex be able to market the developed program effectively. Such limited marketing as could be undertaken at all could only be undertaken by Cyprotex. Thus, Cyprotex would obtain both commercialisation and intellectual property rights in Simcyp that neither Cyprotex nor Sheffield envisaged or expected that it would obtain at the time that the Research Agreement was signed. However, overall, the commercialisation potential of Simcyp would have been severely compromised. Moreover, Cyprotex would gain a considerable advantage over its co-sponsors which was not provided for expressly in the Research Agreement and which none of them could have envisaged or agreed to when entering into the Research Agreement.

It follows that copyright in the Java-based program did not fall within clause 9(d) since that program was not Sponsor IPR covered by clause 9(d) at all.

The Java-based programming work does not fit readily into clause 9(b). That clause applies to “Resulting Intellectual Property” which shall mean:

“individually and collectively all inventions, improvements and/or discoveries whether or not patentable or capable of other intellectual property protection which are conceived and/or made by one or more members or other agents of the University acting either on their own or jointly with one or more employees of the Sponsors in performance of the Programme of Research and relating to its objectives.”

1.

Agency

The first matter to deal with is Sheffield’s suggestion that Cyprotex and Dr Edwards were acting as one of Sheffield’s “other agents” in writing the program so that the resulting program was produced, in the language of clause 9(b), “by one or more members or other agents of the University either on their own or jointly with one or more employees of the Sponsors in performance of the Programme of Research and relating to its objectives.” This agency would have arisen because Sheffield’s use of Dr Edwards’ programming skills enabled it to fulfil its Research Agreement obligation to advertise for and employ a suitable computer programmer to produce the software and that Dr Edwards was, in consequence, “one or more other agents” of Sheffield.

The question of copyright ownership in the Java-based program cannot, however, be resolved by recourse to the principles of agency. Clause 13(e) of the Research Agreement provided that nothing in the agreement should create or be deemed to create a partnership or the relationship of principal and agent between the parties to the Research Agreement. It would not therefore have been possible for Dr Edwards, in his capacity as an employee of Cyprotex, to have acted as an agent of Sheffield for the purposes of clause 9(b) of the Research Agreement, given that Cyprotex was also a party to the Research Agreement and a Sponsor for whom the programming work was being undertaken.

It follows that if Dr Edwards, as an employee of Cyprotex, was also Sheffield’s agent when writing the Simcyp program, the relationship of principal and agent would have arisen between Sheffield and Cyprotex, thereby creating a relationship expressly prohibited by clause 13(e) of the Research Agreement.

The programming work in question cannot readily be said to have been: “conceived and/or made by one or more members or other agents of the University acting either on their own or jointly with one or more employees of the Sponsors”. This is because the program was neither “conceived” nor “made” by Sheffield, given the limited input provided by Dr Rostami. Furthermore, it was neither “conceived” nor “made” by Cyprotex since the idea for the program had already been conceived or made, in a general way, by Sheffield prior to the coming into effect of the Research Agreement when Sheffield devised the project and drafted the content of the Programme of Work contained in the Appendix to the Research Agreement.

It follows that if clause 9(b) is to be applicable, it must be given a strained interpretation whereby Sheffield’s limited role in instructing Dr Edwards and in approving his specifications constituted the making of the program and Cyprotex’s role in writing the program amounted to no more than the making of that program “jointly” with Sheffield.

Alternatively, the Java-based program was, on analysis, excluded from the ambit of clauses 9(b) and 9(c) as well as from clause 9(d) altogether so that clause 9 did not deal with the copyright in that program at all. This latter conclusion would be surprising but understandable since the whole structure of clause 9 was drafted on the basis that the Java-based program would be written by either an employee or an agent of Sheffield yet, in fact, an employee of a Sponsor fulfilled that role.

Given that it would make commercial nonsense of the Research Agreement in general and clause 9 in particular if clause 9 only applied to improvements to the Java-based program and did not apply at all to that program, the words of clause 9(b) should be interpreted so that they include the Java-based program within the ambit of “Resulting Intellectual Property” as defined in clause 9(b). This meaning is one that the parties to the Research Agreement clearly intended and it is a result which can be achieved if the wording of clause 9(b) is given the strained meaning that I have indicated is necessary in order to bring the Java-based program within its scope.

If clause 9 does not cover the Java-based programming work at all, the Research Agreement would be ripe for the implication of a term ascribing copyright in that program to Sheffield. The parties clearly intended that Sheffield should own the copyright in that program and the Research Agreement, particularly its provisions concerning the licences to be granted by Sheffield to the Sponsors and to third parties, only have any practical content if Sheffield owned the copyright in the Java-based program. Moreover, Sheffield would be unable to market the Simcyp program once it had been developed or to sell the marketing rights in that program unless it owned the copyright. These considerations suggest that a term should be implied into the Research Agreement to the effect that the copyright in the Javabased program should vest in Sheffield if that was not already the consequence of the express wording of clause 9.

In support of the argument that such a term should be implied, Sheffield placed reliance on this passage from the Ray v Classic FM case:

“(7)

Circumstances may exist when the necessity for an assignment of copyright may be established. As [counsel for the plaintiff] has submitted, these circumstances are, however, only likely to arise if the Client needs in addition to the right to use the copyright works the right to exclude the Contractor from using the work and the ability to use the copyright against third parties. Examples of when this situation may arise include:

(a)

where the purpose in commissioning the work is for the Client to multiply and sell copies in the market for which the work was created free from the sale of copies in competition with the Client by the Contractor or third parties;

(b)

where the Contractor creates a work which is derivative from a pre-existing work of the Client, eg when a draughtsman is engaged to turn designs of an article in sketch form by the Client into formal manufacturing drawings and the draughtsman could not use the drawings himself without infringing the underlying rights of the Client;

(c)

where the Contractor is engaged as part of a team with employees of the Client toproduce a composite or joint work and he is unable, or cannot have been intended to be able, to exploit for his own benefit the joint work or indeed any distinct contribution of his own created in the course of his engagement: see Nichols Advanced Vehicle Systems Inc v Rees [1979] RPC 127 at 139 and consider Bogrich v Shape Machines unreported 4 November 1994 and in particular page 15 of the transcript of the judgment of Aldous J.”

Transposing these Ray v Classic FM principles to the facts of this case, I would venture a fourth example of where copyright should, by implication, vest in the Client, being a person in Sheffield’s position:

“where the purpose in commissioning the work was to transform a research and development program containing a ground-breaking new technique for drug research that had been developed by the Client into a user-friendly program capable of being marketed to the pharmaceutical industry world-wide on a major scale by the Client and where the Contractor creating that marketable program would already have been granted by the commissioning agreement a licence to use the program for its own internal purposes connected with different but related areas of research but had not been granted the right to market the program without a subsequent agreement being negotiated with the Client.”

Thus, I accept that clause 9 of the Research Agreement should be read as if it was subject to an implied term that the copyright in the Java-based program should vest in Sheffield.

For all these reasons, I conclude that either:

Six further considerations need to be dealt with.

Firstly, since Sheffield is not actually using the Java program but has developed its own C ++ program, it might be considered that any declaration as to its rights to the copyright in the Java-based program is purely academic. However, it is clear from the evidence of Professor Tucker that the Java-based program retains potential use and practical utility and Sheffield is entitled to protection in, and a delivery up of, the program, albeit that Cyprotex may retain a copy for its own internal purposes given the terms of the licence granted to it by clauses 9(e) and 9(g) of the Research Agreement.

Secondly, Sheffield does not seek directly a declaration of its entitlement to its copyright in the algorithms, data bases and MathCAD program it developed as Background IPR. However, since copyright in the Java-based program was in issue and since that program is an improvement of, and a development out of, the earlier work, it would seem appropriate to grant Sheffield a comprehensive declaration of its rights in Simcyp and in the MathCAD program and any other of its forebears and antecedents that are also subject to Sheffield’s rights of copyright.

Thirdly, Sheffield does not seek an injunction to restrain any infringement of its copyright in Background IPR since no such threat has ever been intimated.

Fourthly, Dr Edwards was the author of the specifications. However, these were produced by an employee of Sheffield jointly with Dr Edwards so that they are subject to clauses 9(b) and 9(c) of the Research Agreement. Alternatively, they were not subject to any part of clause 9 50 that they would be subject to the same implied term as that governing the copyright in the Simcyp Java-based program. It follows that these specifications should be subject to the same declaration in favour of Sheffield as governs the Simcyp Java-based program itself.

Fifthly, Sheffield seeks orders for the assignment and the delivery up of the copyright and all disks and other material containing the Java-based Simcyp program. Although Sheffield is entitled to an assignment of the copyright and to the originals of all documents, files, discs or tapes containing all or any part of the Simcyp program, Cyprotex is entitled to a wide-ranging licence to use the program for its own research purposes and is therefore entitled to retain copies of all this material currently in its possession.

Sixthly, Sheffield seeks an enquiry as to damages. It seems unlikely that Sheffield has incurred any additional damages so that, at first sight, an enquiry would appear to be unnecessary and the court ought not to consider ordering one.

With these considerations in mind, Sheffield would not appear to be entitled to an order requiring Cyprotex to deliver up all copies of Simcyp materials in its possession nor to one providing for an enquiry as to damages. I will, however, allow the parties an opportunity to make submissions on these matters before the order giving effect to this judgment is perfected.

For all these reasons, Sheffield is entitled to the following declaration:

“The copyright and any other intellectual property:

(1)

in or over and any works relating to Simcyp created by Dr Edwards or any other servant or agent of Cyprotex Discovery Limited or Medeval Limited; and

(2)

in or over MathCAD and any algorithm or data base used in its compilation; and

(3)

in any specification created by or with the assistance of Dr Edwards, Dr Rostami or Professor Tucker relating to Simcyp or MathCAD,

is owned by The University of Sheffield.”

The draft judgment is confidential to the parties until handed down in open court but it may be shown to counsel, solicitors and anyone within the parties’ respective organisations who has a reasonable need to be consulted about or to know the contents of this draft judgment.

The judgment will be handed down at a robed hearing on Friday 21 February at 9.30 am. If either party wishes to move this date or time, a joint application should be made to my clerk, Anne Farrelly, which should be accompanied by alternative mutually convenient suggested dates and times.

At the handing down hearing, I will deal with: