Chalmers Bioscience Venture

--------------------------------------------------------------------------------

GENERAL BACKGROUND

The understanding of molecular processes in biology and medicine, and the interpretation of data from large scale experiments, require new and more powerful tools from advanced computer science, mathematics, and statistics. These tools are used to build data bases, construct search and learning algorithms, in optimisation and modelling, and to draw statistical conclusions from often highly complex and noisy data. Bioinformatics in this broad sense is rapidly becoming an essential tool in biological and medical science. Computerised sequence analysis for gene finding, gene clustering, expression analysis, etc., are relatively novel and promising subjects and interest in bioinformatics is increasing rapidly.

SPECIFIC BACKGROUND

In winter 1998, the Swedish Foundation of Strategic Research (SSF) asked Swedish universities if they were willing to host a national centre of bioinformatics, partly financed by SSF. Chalmers University of Technology (Chalmers) and Gothenburg University (GU) entered this competition. Anders Blomberg, Dept. of Cell and Molecular Biology, GU, lead a team of cooperating scientists representing several pertinent disciplines in the application process. This effort did not result in a center located in Gothenburg,, but later on 7.5 million SEK were granted from SSF for the start of a Bioinformatics Research Programme. The final version of the SSF application, dated January 1999 (unabridged as well as short form) can be found on:

See specifically the descriptions of the biological platforms in relation to experimental disciplines, existing data bases, possible partners in industry, etc.

In June 1999, bioinformatics was singled out as a necessary core facility needed for Chalmers bioscience research initiative, and several activities within the programme started this academic year as a collaboration between the School of Mathematics and Computer Sciences (MD) and several departments of biology, medicine and chemistry at GU. Chalmers has decided to support the programme with 5.9 MSEK for this academic year, which includes overhead costs. This decision precipitated the above mentioned SSF support. The Stochastic Centre, a separate research unit at MD is responsible for the financial administration of the programme.

The bioinformatics programme is lead by a steering committee consisting of Peter Jagers (chairman) and Mats Rudemo, both Professors of Mathematical Statistics at Chalmers, Gunnar Bjursell, Professor of Molecular Biology at GU, and Lena Gustafsson, Professor of Molecular Biotechnology at Chalmers.

Members of the newly formed international advisory board are Graham Cameron, CBI Cambridge; Stephen Oliver, University of Manchester; Mattias Mann, University of Odense; Chris Sander, Millennium Predictive Medicine and MIT Genome Center; Mike Waterman, University of Southern California; and Elisabeth deTurckheim, INRA, Paris. We are extremely satisfied with the composition of this board, representing a wide range of expertise and disciplines.

Olle Nerman, Professor of Mathematical Statistics at Chalmers and Anders Blomberg, Associate Professor of Microbiology work together almost full time as academic leaders of the venture.

In brief, the following activities are being initiated:

Paired PhD projects (start in Jan. 2000); an international master's programme in bioinformatics (starts in Sept. 2000); a thematic research semester in the spring of 2000, with invited leading scientists; shared space for joint work; and joint bioinformatics seminars. Currently, funding is mainly received from Chalmers, while GU covers rent and some computer equipment. The thematic research semester is also supported by the Gothenburg Stochastic Centre and the NTM-programme, a national programme in applied mathematics supported by SSF for 1997 – 2001. The previously mentioned SSF funding of 7.5 MSEK is reserved for a strategic international recruitment of a senior scientist (vide infra).

Another event of major importance for the bioinformatics activities in Gothenburg is the joint effort by the University of Lund/Malmö, Chalmers and GU in the area of functional genomics, "Swegene" (Postgenomic Research and Technology Programme in South Western Sweden). The main ingredient in this programme is an application to the Knut and Alice Wallenberg Foundation (KAW) covering quite extensive activities over a period of five years.

Discussions are under way with one of Sweden’s multinational pharmaceutical companies concerning industry support of the bioinformatics research programme.

GENERAL GOALS

The steering committee for this new interdisciplinary effort has formulated a set of long term goals for research, graduate education, and the operational format of the initiative. They are:

• To build up a broad competence in bioinformatics.
• To reach an international top position in the second wave of bioinformatics, such as expression analysis, mathematical modelling of signalling networks and other cell systems
• To build a strong group in epidemiological genetic statistics.
• To establish strong research in areas of computer science that relate to bioinformatics, including biological algorithms, databases, and visualisation.

• To build bridges between biology/medicine and mathematics/statistics/computer science.
• To strengthen industrial research collaboration.
• To strengthen international research collaboration.

ACTIVITIES AND PLANS

A key ingredient in our strategy are the graduate student pair-projects. An inventory procedure was initiated, in which most potentially interested academic researchers in biology and medicine in Gothenburg were contacted via e-mail, with a "Call for Proposal", see:

The result was 11 potential pair PhD-projects out of which 6 were selected by the steering committee based on scientific merits, realistic potential, and importance for the overall programme. The international advisory board has evaluated the scientific merits of the 6 projects chosen in this first step. Several of the projects that were not selected in this first round were, however, judged by the steering committee to be of potential interest for future recruitment of graduate students, and they may also result in future senior collaborative research efforts. Full project descriptions can be found at:

Supervisors: Olle Nerman, School of Mathematical and Computing Sciences, Mathematical Statistics, Chalmers.

Anders Blomberg, Dept. of Cell and Molecular Biology, Microbiology, GU.

Supervisors: Mikael Kubista and Tomas Simonsson, Dept. of Chemistry, Molecular Biotechnology, Chalmers.

Olle Nerman and Serik Sagitov, School of Mathematical and Computing Sciences, Mathematical Statistics, Chalmers.

Olle Nerman, School of Mathematical and Computing Sciences, Mathematical Statistics, Chalmers.

Supervisors: Lena Gustafsson, Dept. of Chemistry, Molecular Biotechnology, Chalmers.

Carl Johan Franzén, Dept. of Chemistry, Chemical Reaction Engineering, Chalmers.

Bernt Wennberg, School of Mathematical and Computing Sciences, Mathematics, Chalmers.

Supervisors: Tore Samuelsson and Per Elias, Dept. of Medical Biochemistry, GU,

Dag Wedelin, School of Mathematical and Computing Sciences, Computing Sciences, Chalmers & GU,

Supervisors: Stefan Hohmann, Cell and Molecular Biology/Microbiology, GU.

Per Sunnerhagen, Cell and Molecular Biology/Molecular Biology, GU.

Dag Wedelin, School of Mathematical and Computing Sciences, Computing Sciences, Chalmers & GU.

The projects attracted close to 60 applicants. We have just interviewed a dozen highly qualified candidates, and the research activities will start in January 2000.

Most certainly, all 6 projects will be launched, allthough some projects will be carried out by a single graduate student and not a pair as planned. This has been made possible through external funding of parts of the projects 2, 4 and 5.

Our intention is that these graduate students will be seen and treated as "bioinformatics students", with identities and roles different from the traditional ones in their respective discipline. The collaboration part of their research will mostly take take place in the afore mentioned shared space, which is located at the campus of the Schools of Natural Science and Medicine, GU, close to the Chalmers Campus. Many of the planned study groups and seminars will also take place there.

Currently, a major weak point in our bioinformatics initiative is the low level of involvement by computer scientists. A first step to improve this situation was recently taken by advertising a tenure track Assistant Professor position in Computer Science/Bioinformatics, see:

Furthermore, we will most probably use the SSF-grant for recruiting an additional Associate Professor or a Professor of Computer Science/Bioinformatics already in the spring of 2000. We believe that the approaching Master Programme, and the ongoing Thematic Semester and pair PhD-projects will be attractors of major importance in the recruiting process.

A thematic semester devoted to bioinformatics and Genetical Statistics will be organised by the Gothenburg Stochastic Centre during spring of 2000. Professor Michael Waterman, University of Southern California, Los Angeles, has been appointed Chalmers Jubilee Professor for spring 2000, and will spend February, March and May in Gothenburg. Up to ten international and ten Swedish guests are expected to spend longer periods (3-8 weeks) at the Stochastic Centre during this period.

A national course for graduate students with a background in mathematics, statistics or computer science will be given in Uppsala and Gothenburg (2´ 3 days) in February and March. The course was initiated by our programme and Stochastic Centre and will be a joint activity between us and the recently founded bioinformatics centres in Uppsala and Stockholm. For more information see:

Furthermore, as a part of the thematic semester, a workshop with 50-60 participants will be held in Gothenburg on May 9-13. A special open session is planned on May 12 with speakers from the biological/medical community.

Chalmers has just decided to support an initiative from the Bioinformatics Programme to start an international Master Programme in Bioinformatics in September 2000. Marita Olsson, Mathematical Statistics, Chalmers, Tore Samuelsson Medical Biochemistry, GU, and Dag Wedelin, Computing Sciences, Chalmers, have prepared the detailed curriculum for a 1.5 year long programme, to which a steady stream of both "bio" and "math" students are expected (hopefully 30 per year). More information is available on:

The Master's Programme is expected to have several important effects on the research in bioinformatics: First, recently engaged senior scientists will over the first years get training through participating in and following the courses. Second, the courses will be used extensively for training the recruited graduate students. Third, the programme will generate a lot of tutorial work that will eventually be performed by the graduate students in the bioinformatics programme. Fourth, we believe that the programme will generate a steady in-flow of good PhD-student candidates in future years.

Several existing educational programmes at Chalmers will benefit from some of the courses in this international master's programe, mainly the Biotechnology, Technical Mathematics, and Computer Engineering Programmes. By taking courses in the Bioinformatics Master's Programme, such students will also be better prepared for a future career in bioinformatics .

Two popular courses related to the master's programme have been developed this autumn; a graduate course in "Statistical Methods for Gene Localisation in Complex Diseases" offered at the medical faculty and an introductory bioinformatics course offered to advanced undergraduate students at Chalmers and graduate students in biology and medicine at GU.

Since the beginning of 1999 a regular series of seminars in bioinformatics has been held (approx. 10 seminars / semester) to the research community at Chalmers and GU.

In the autumn 1999 we started of a joint working group in sequence-related bioinformatics with broad participation among several different groups from both mathematics/computer science and biology/medicine.

The Department of Mathematical Statistics and the Stochastic Centre have already been engaged in epdemiological genetic statsitics for a couple of years. Four graduate students at the centre work with linkage and association research. Two of the students are connected to gene-scan projects at the Department of Clinical Genetics, GU, studying Coeliac disease, Psoriasis and Multiple Sclerosis, and will receive their licentiate degrees this academic year (Dec. 1999 and Feb. 2000). The third student works part time at Astra-Zeneca., and is there engaged in a heart-related genome scan. The fourth graduate student was initially engaged in methods for construction of different genetic maps (in collaboration with Genetics, GU) and recently started studies in quantitative association techniques (based on ideas of William Amos, Cambridge). The genetical statistical efforts can be developed in several directions in concert with the rapid technological developments and possibilities in the field. The way genes related to complex diseases are identified will possibly change drastically when the precision in expression analysis improves and all sorts of combination methodology will be tried. Obviously, the need of well-trained statisticians in the field is and will be large.

Several undergraduate and two graduate projects, run in collaboration with the Swedish pharmaceutical industry (Astra-Zeneca) and a visualisation/bioinformatics company (Spotfire), are connected to the programme. In the Astra-Zeneca collaboration, involving one of the graduate students, much effort is spent trying to understand expression data generated from Affymetrix-type DNA chips. The second graduate student (who was also mentioned in the previous paragraph) works on sibpair techniques for multi locus gene localisation (based on a suggestion of Terry Speed, Berkeley).

Scientists at Mathematical Statistics have very close relations to several statistically oriented bioinformatics groups in the US, and we firmly believe that the thematic spring semester in 2000 will strengthen these contacts as well as create new ones.

IMPLEMENTATION

We wish to keep the current organisational format with a steering committee and two academic leaders/"executive officers" for at least another two years. In the future, the Swegene Initiative (especially in case it is funded by the Wallenberg Foundation) and the planned employing of high class senior bioinformaticians will certainly require reevaluation of the programme organisation. A future possibility of inviting industry representatives will be considered.

1. At least 10 graduate students supported by the programme (80% research, 20% teaching). 6-7 of these are being recruited in Dec. 1999. The total cost of these ten students is roughly 5 MSEK/year. (This figure may turn out to be somewhat optimistic, depending on Chalmers and GU policy decisions)
2. Support for one biologist (today, Anders Blomberg) and one computer scientist/mathematician / statistician (today, Olle Nerman), the steering commitee, the advisory board, etc., with at least 1 MSEK/year.
3. Support for bioinformatics research by younger associate professors in mathematics/ computer science and mathematical statistics that engage in the development of the international master's programme (the goal is to get specific key bioinformatic competences needed for a high quality master's programme) 0.6 MSEK/year.
4. The corresponding support of researchers in the international master's programme on the biological/ medical side will require 0.4 MSEK/year.
5. To hire and support two tenure track assistant professors in computer science/bioinformatics on a half time basis (50% will be teaching at the Master's Programme). Cost: roughly 0.7 MSEK/year.
6. A yearly sum of 0.6 MSEK will be reserved for computer costs, hardware, software, and maintainance.
7. For joint research projects in collaboration with other special ventures sponsored by Chalmers, and for support or collaboration with other strong bioinformatics research projects at Chalmers or GU we wish to reserve 1 MSEK/year. Towards the end of this application we have explicitly mentioned a number of names of interesting researchers that we wish to continue discussions with concerning such support and/or collaborations. Our special interest in this context are researchers involved in the programmes "From Gene to Product" and "Biomolecular Structure and Function", some of whom we already collaborate with on the pair Ph.D. projects. which we already collaborates with in the Phd pairs. Furthermore, we are sure that there are several others whom we have not yet identified and believe that more will turn up in the future
8. Two flexible postdoc positions, which will be used either to support the most successful pair projects or initiating new collaborative work. Cost: roughly 0.7 MSEK/year.

* In addition to this budget we would like to use the SSF-grant of 7.5 MSEK to support the recruiting of a Bioinformatics oriented Computer Science Professor or Associate Professor. The money would be allocated to pay this cost 4-5 years as well as the costs of a PhD or Postdoc working with the Professor.

*For the cost of necessary international collaborations, guest professors, sabbaticals, conferences etc, we hope to be able to raise money through the usual channels.

*In the Swegene application to the Knut and Alice Wallenberg foundation (see above) we ask for support for the earlier initiated genetic epidemiology statistics and for support of initiation of biological database related research. We also ask for support for developing a strong expression analysis research. The outcome of that application is not yet decided. This makes the balance between different research groups in the present application rather hard to fix. For example, we think that it is natural to start at least one PhD-pairproject in the genetic epidemiology field even if the Wallenberg Application is turned down. Thus, we wish to postpone the decisions, e.g. concerning what PhD-projects should be initiated next to the steering committe and the advisory board.

*We will also try to raise further external money to be used to support one more senior bioinformatics position to be announced in spring 2000 along with the one sponsored by SSF. We believe that the thematic semester effort, the pair-projects, the Master’s Programme and the enhusiasm of already engaged scientists will make it possible to attract really competent bioinformaticians.

*Some of the money received from Chalmers in June1999 have been allocated for support of the thematic semester in spring 2000. This money is also intended to cover the development cost for the master’s programme. We have decided to leave both these costs out in the budget above, and wish to discuss these matters further at a later point in time.

We have chosen to describe only the changes compared to the first two years. Numbers below refer to items in the previous paragraphs.

A gradual expansion of external funding is expected for these years. We will try hard to compete for resources from national and European funds for PhD-projects, postdocs and for senior research. We shall keep the strategy to primarily support interdisciplinary projects. If the International Master's Programme works out as well as anticipated, we expect a quality-controlled growth of the Bioinformatics Programme.

2. We expect that the organisational administrative difficulties, that cannot be avoided during the first and second years will gradually disappear and that we will need somewhat less extra money for that purpose over the next three years (certainly usual overheads will still be needed). We have therefore decreased this part of the budget to 0.5 MSEK/year.

3 and 4. The costs for initiating bioinformatics research among senior researchers involved in the International Masters Programme will be cut to 0.5 MSEK/year. This is motivated by the anticipation that external funding will by then be coming in to the programme.

Consequently, we would like to move 1 MSEK/year to promoting collaborations and supporting the bioinformatics research mentioned in paragraph 6 above. These activities then get a yearly budget of 2 MSEK. This will hopefully result in a natural and controlled growth of external research activites.

Furthermore, we shall make a strong effort to receive national and European funding for each of these projects.

PhD students 5.0 MSEK

Project management, supervision, etc. 1.0 MSEK

Research for teachers in Master's Programme 1.0 MSEK

Computer costs 0.6 MSEK

Research for 2 Comp. Sci. Associate Profs. 0.7 MSEK

External research 1.0 MSEK

Postdocs 0.7 MSEK

--------------------------------------------------------------------------------

Sum per year 10 MSEK

PhD-students 5.0 MSEK

Project management, supervision, etc. 0.5 MSEK

Research for teachers in Master's Programme 0.5 MSEK

Computer costs 0.6 MSEK

Research for two Comp. Sci. Associate Professors 0.7 MSEK

External research 2.0 MSEK

Postdocs 0.7 MSEK

--------------------------------------------------------------------------------

Sum per year 10 MSEK

PLATFORMS: COMPETENCES AND POTENTIAL COLLABORATORS

For a concise inventory of the platforms that make us believe that we can achieve our goals, we wish to refer to the earlier application to SSF. Here we will give a brief list of primarily involved departments and individuals, and indicate some relevant references. Developing a large new research environment on a collaborative basis is certainly not simple and guaranteeing full success in all projects and at all steps of such a process would be ridiculous. Thus we believe that an active and flexible steering committee and advisory board is of utmost importance, so that necessary changes of plans can be made without delays.

Four units on the biological and medical side are primarily involved in this application: the departments of (i) Cell and Molecular Biology at GU, (ii) Molecular Biotechnology at Chalmers, (iii) Medical Biochemistry at GU, and (iv) Clinical Genetics at GU. Detailed descriptions of the expertise and research activities at all these units, except for the Deptartment of Clinical Genetics, are given in the proposal "From Gene to Product" coordinated by Lena Gustafsson. Thus, we will in the former three cases only give brief descriptions of some of the already involved researchers:

Anders Blomberg is an Associate Professor in microbiology at GU, and is currently acting as academic leader/executive officer together with Olle Nerman (Mathematical Statistics) in the initiated bioinformatics programme. Anders Blomberg’s main research interest is in the understanding of the molecular physiology of stress responses in the eukaryotic model system yeast (Saccharomyses cervisiae). These analyses are focused on i) large scale expression analysis of proteins (proteomics), ii) phenotypic analysis of deletion mutants, and iii) molecular mechanisms involved in stress induced transcriptional activation. He is currently involved in a large EU-funded project for the functional characterisation of genes utilising proteomics tools for analysis of deletion mutants. Anders Blomberg is also maintaining a web-based yeast proteomics database.

Tore Samuelsson, Associate Professor in Medical Biochemistry, has since 1994 been responsible for bioinformatics courses as well as a sequence analysis service facility at GU. His background is in protein and nucleic acid biochemistry. Current bioinformatics projects include analysis of sequence data related to the signal recognition particle, and to the mediator subunit of RNA polymerase.

Per Sunnerhagen is an Associate Professor in Molecular Biology, GU. His research focuses on intracellular signalling in yeast. Recently, genomic and integrative approaches are being used to elucidate the network of interactions that make up regulatory pathways.

Fredrik Ståhl is an Associate Professor in Genetics at the Dept. of Cell and Molecular Biology. He has a background in molecular genetics and his dissertation dealt with gene amplification of mouse tumour cell lines. During the last years, Fredrik Ståhl has been responsible for the development and maintenance of the rat genome database initiated by Professor Göran Levan, RATMAP (http://ratmap.gen.gu.se).

Mikael Kubista, Professor in Biotecnology, has an interest in in vitro DNA technolgy. He is also working with multivariate data anlysis and have extensive experience in method development to

analyse multidimensional data.

Department of Clinical Genetics has in the 1990’s transformed from being mainly a clinical hospital unit to a modern human genetic research unit within Gothenburg University. Nowadays it is one of the major Swedish research centres for Epidemiological Genetics. The department is lead by Associate Professor Jan Wahlström. Several large genome scans of family data are underway, or will soon be initiated (Coeliac disease, Multiple Sclerosis), and a big gene hunting project focusing on Psoriasis genes has been successfully performed as a joint industrial venture together with Glaxo Wellcome. Scientific leaders of the research in molecular genetics are a geneticist, Associate Professor Tommy Martinsson, and a molecular biologist, Lena Samuelsson, Ph.D..

Below find six references of work performed by some of the involved biological and medical researchers and/or their groups (others can easily be found in the descriptions of the pair PhD projects on the Bioinformatics home pages).

Polevoda, B., Norbeck, J., Takakura, H., Blomberg, A., and Sherman, F. (1999) Identification and specificities of N-terminal acetyltransferases from Saccharomyes cerevisiae. The EMBO Journal, 18: 6155-6168

Samuelsson, T. and Zwieb, C. (1999) Nucleic Acids Res 27:169-70. The Signal Recognition Particle Database (SRPDB)

Levan, G., Ståhl, F., Klinga-Levan, K., Szpirer, J. and Szpirer, C. (1998) The rat gene map. ILAR Journal 39:132-137

Bilsland E., Dahlen M., and Sunnerhagen P. (1998). Genomic disruption of six budding yeast genes gives one drastic example of phenotype strain-dependence. Yeast 14, 655-664.

Kubista M., Nygren J., Elbergali A., and Sjöback R. (1999) Eliminating the need for reference samples. Critical Reviews in Analytical Chemistry 29, 1-28.

Enlund F., Samuelsson, L., Enerbäck, C., Inerot, A., Wahlström, J., Yhr, M., Torinsson, Å., Riley, J., Swanbeck, G. and Martinsson, T. (1999) Psoriasis susceptibility locus in chromosome region 3q2 identified in patients from southwest Sweden. European Journal of Human Genetics 7 (7):783-790

The School of Mathematical and Computing Sciences is a part of both Chalmers and GU. The activities of the two universities are highly integrated and have been held up as a model in national evaluations. Within the school, research and doctoral and undergraduate studies are offered in the disciplines of mathematics, applied mathematics, industrial mathematics, mathematical statistics and computing science to an extent and with a range which is outstanding compared to other Swedish universities. The school also offers a specialization in engineering mathematics. The department has approximately 270 employees, of which 140 are researchers and teachers, 95 are doctoral students, and 35 administrative and technical personnel. Each year, about 1500 students from GU and 2800 students from Chalmers study here.

Since autumn of 1996, Chalmers University of Technology and the Swedish Foundation for Strategic Research through the National Network in Applied Mathematics, NTM, support this international research centre in mathematical statistics at Chalmers and GU. Internationally recognised researchers as well as younger researchers from both Sweden and outside of Sweden are affiliated with the centre, often in during periods devoted to particular themes. A central goal of the centre is to strengthen collaboration with Swedish industry by having researchers from industry participating in the centre's activities as well as developing joint projects between industry and academia..

The Stochastic Centre is governed by a council from the department of Mathematical Statistics, Chalmers and GU. The council consists of Sture Holm, Olle Häggström, Peter Jagers, Jacques

deMaré, Olle Nerman, Holger Rootzén (chair) and Mats Rudemo (director).

The research at the department of Mathematical Statistics and at the Stochastic Centre includes medical and biological statistics, population models, statistical extreme value theory, reliability theory, quality control, sampling theory, particle systems, percolation, image and signal analysis, Markov and stationary processes, as well as statistical modelling of material fatigue. Research projects are often conducted in close collaboration with scientists from other disciplines and industry.

In addition to organising visitors programmes, the Stochastic Centre supports PhD students, post-doctoral researchers and research of senior scientists.

The mathematics department has extensive research in many areas of relevance for this application. Some examples are: numerical methods for solution of differential equations and large sparse linear equations, and a rapidly evolving optimisation group. There is also a strong group specialising in theoretical aspects of Boltzmann equations (from which Bernt Wennberg originates, see below). The potential for recruiting bright young mathematicians into the bioinformatics field is great.

The computer science department hosts strong research in the fields of algorithms, distributed computing systems, cognition technology, formal methods, computing systems, functional programming, concurrency and programming logic. The two most relevant research groups for the Bioinformatics Programme are the algorithmic group and the cognition group. The algorithmic group focusses on discrete optimisation, machine learning and since approximately one year bioinformatics. The visualisation system "Spotfire" has its roots in research performed in the cognition group. Spotfire is extensively used in the pharmaceutical industry in various bioinformatics contexts.

Olle Nerman, coordinator of the present application is Professor in Mathematical Statistics at Chalmers. He started his research in branching processes and probabilistic population dynamic models and has in the 90’s broadened his research activities in several different directions. In addition to recent interest in genetical statistics and bioinformatics, biostatistics and survey sampling are main areas of interest. Nerman has supervised more than ten students to a licentiate or doctoral degree, mainly in population dynamics, biostatistics and survey sampling, but his students has also worked on topics as varied as stochastic geometry, mathematical analysis of epidemic spread models, and Markov modelling of reliability in airplane motorsystems.

Dag Wedelin, Associate Professor in Computer Science, has a broad interest in discrete optimisation, machine learning and algorithms. He was the initiator and coordinator of a recently completed EU-funded international project on airline crew scheduling optimisation techniques and their computer implementations. He has also worked with efficient algorithms for multivariate data analysis using

Graphical models (Markov graphs).

Peter Jagers, Professor of Mathematical Statistics and the present chairman of the Steering Committee of the Bioinformatics Programme has a long standing interest in population dynamics, especially branching processes. Mathematical and statistical modelling of the relation between micro-level phenomena, e.g. cell cycles and population compositions, and population size dependent branching processes are two examples.

Mats Rudemo, Professor of Mathematical Statistics and director of the Stochastic Centre, has a broad research interest in statistics in biology and engineering including image analysis, spatial statistics, , as well as precision forestry.

The Associate Professors of Mathematical Statistics Ziad Taib and Serik Sagitov have backgrounds in branching processes. Ziad Taib’s research is already biologically oriented (e.g. cell cycle models, neutral genetic evolution in branching models), and the potential contributions from Serik Sagitov’s recent research on the probabilistic side of coalescent processes have a good potential not only in evolution modelling, but also in the statistical phylogenics area.

Assistant Professor Marita Olsson is the coordinator of Chalmers’ new Bioinformatics Master’s Programme. She has experience of partially observed Markov processes and is currently involved in the human gene finding activity.

Associate Professor Marianne Månsson is interested in Stochastic Geometry and Poisson approximation of various complicated dependent stochastic events. She has studied certain sequence matching asymptotics of relevance in genefinding and clustering in connection to string comparisons.

Associate Professor Bernt Wennberg has a background in Boltzmann equation research and is the head of the mathematical consulting group at the department.

Below find nine references from the researchers involved in the mathematics, statistics and computing science:

Nerman, O. and Jagers, P. (1996) The asymptotic composition of supercritical multi-type branching populations. Springer Lecture Notes in Mathematics, 1626, 40 - 54 (1996).

Åhlström, L., Olsson, M., Nerman, O. (1999) A Parametric Estimation Procedure for Relapse Time Distributions. Lifetime Data Analysis 5, 113-132.

Wedelin, D. (1996). Efficient estimation and model selection in large graphical models. Statistics and Computing 6, 313-323.

Larsen, M. and Rudemo, M. (1998) Optimizing templates for finding trees in aerial photographs, Pattern Recognition Letters 19, 1153-1162.

Taib Z., (1997) A branching process model for the mutation-selection balance. Advances in Population Dynamics- Molecules Cells and Man. O Arino, D Axelrod and M. Kimmel Eds. World Scientific Publishing.

Sagitov S., (1999) The General coalescent with asynchronous mergers of ancestral lines. J. Appl. Prob. 36, vol 4.

Wennberg B.,(1999) An example of non-uniqueness for solutions of the Boltzmann equation, J. Stat. Phys. 95 no 1 / 2 , 473-481.

Månsson M.,(1999) On compound Poisson approximation sequence matching. Preprint, the Department of Mathematical Statistics, Chalmers University of Technology and Göteborg University

Nilsson S. Two contributions to Genetic Linkage Analysis (1999). Licentiate thesis at the Department of Mathematical Statistics, Chalmers University of Technology and Göteborg University

There are obviously a great number of collaboration possibilities with the other Chalmers programmes. The "Biomolecular Structure and Function" programme (coordinator Per Elias) and the "From Gene to Product" programme (coordinator Lena Gustafsson) are of course of great interest, as is reflected e.g. in our paired Ph.D. projects. An area of research that we have not included explicitly

in this application is statistical analysis of biological image data which is Mats Rudemo's speciality. In this field the "Bioengineering" programme (coordinated by Mats Viberg) has potential collaborators.

We believe that it would be interesting to establish some form of collaboration with Mats Nordahl and some of the students in the Complex System Group at the Theoretical Physics Department. He has a long biological modelling and bioinformatics interest and also works with machine learning, neural nets and genetical algorithms.

The young Assistant Professor Mattias Wahde, works in the Mecatronics group at the School of Mechanics. Mattias has recently held a postdoc position at Nordita in Copenhagen and there started collaborative research in neural network modelling of time series of expression data together with Professor John Hertz.

The biostatistics group in Mathematical Statistics, lead by Professor Sture Holm, is a natural partner in many general statistical inference problems in connection to both genetical statistics and bioinformatics. As an example, so called penetration probabilities, extensively used in statistical modelling of inherited diseases, will commonly have to be age-dependent so that many methods from survival analysis are highly relevant in this context. Techniques to study interaction of genes and drugs and/or environments are also closely related to analysis of variance methods etc.. We also look forward to all the exciting diagnostic possibilities that expression analysis profiles will facilitate. Many techniques, already known in biostatistics are likely to be central in this area.

We specially wish to mention the possibility to integrate statistical and computational methods in phylogenics research as an important natural activity of the programme. Evolution and function are certainly very tightly coupled in bioinformatics and the understanding of both is necessary in much bioinformatics. There is a strong phylogenics group at the Zoology department lead by Associate Professor Per Sundberg. We will try to establish some form of collaboration with this group, at a level depending on our human resources and funding. There are also potential, not yet contacted or even identified, collaborators in phylogenics in the Botanical Institute, GU.

We have already established good relations to both the Stockholm Bioinformatics Centre and the Linné Centre in Uppsala and we will arrange a national graduate course together in February-March 2000. We wish to pursue future joint projects and graduate course arrangements with these centres.

Through the Swegene initiative it is likely that our links with Bioinformatics researchers in Lund will grow even stronger than they are today. Already, the Stochastic Centre have extensive collaboration with mathematical statisticians in Lund in several fields, e.g. telecommunication and fatigue statistics.

Traditionally, Mathematical Statistics has a close contact with the statisticians at ASTRA-ZENECA including both clinical and preclinical research. Recently a joint project has been started also in drug delivery research supported by the Swedish National Board for Technical Development, TFR. We will try to establish as many bioinformatics and genetical statistics research collaborations as possible, especially at the master’s and doctoral student level. Similarly, we intend to establish new research collaboration with Spotfire in the visualisation area.

Finally, we would like to emphasise the enthusiastic spirit prevalent among all scientists involved in the Bioinformatics Programme. We hope that this enthusiasm will stay with us, and that it will be transmitted to all the young scientists that we are currently recruiting.

Gothenburg 1999-12-21

For the Bioinformatics Programme

Olle Nerman

Programme Application Cordinator and Professor in Mathematical Statistics at Chalmers