The ISBGroup Blog

2011 was an eventful year for us, when many things happened. Here are a few of the main highlights:

* In June, we arranged a series of three workshops on systems biology: on conclusions despite uncertainties, on multi-level modelling of type 2 diabetes, and an internal one regarding systems biology at Linköping University

* We had quite a few students who finished their Master Theses: Robert Palmér, Mikael Forsgren, Ulrike Munzner, Fianne Sips, Karin Lundengård, and Linnea Järvstråt. All of these are now working with systems biology at various places out in the world.

* The new full-time professor Mikael Benson has started. He is a professor in clinical systems biology.

* Our two guest-professors Jens Timmer and Hiroaki Kitano has started their employments

* We attended the ICSB (International Conference on Systems Biology, in Heidelberg) with a large attendence. We had some 10 posters, and equally many participants, from Linköping; Gunnar Cedersund gave a presentation; and Mikael Benson arranged a spin-off event.

* In Dec of 2011 we arranged a new workshop. This time with the themes: Hierarchical modelling, Decision-support in clinical practice, Type 2 diabetes, and Liver diseases.

* On multi-level modelling of type 2 diabetes, we also got published one of our most important papers yet: describing the linking between intracellular mechanisms and whole-body glucose homeostasis.

* I was awarded a VR-FoAss, which is a position financed by the Swedish Research Council. It is a relatively high-status position (<8% success rate), and it allows me to act a little bit more independently of my joint group with Peter Strålfors. I have therefore started the process of dividing away part of my time to IMT (currently 40%), the neighbouring department on biomedical engineering. IMT is situated a few meters away from my IKE group with Peter, i.e. still at the hospital campus (HU), but it is formally located within the technical faculty, and is mostly populated by engineers. From here, I can and plan to start building up a more independent modelling environment.

One of the major tasks this spring is to at last start forming a core-facility in systems biology and mathematical modelling.

This will happen as a natural out-growth from a process that has happened within the ISBgroup during the last 2 years or so. During this time we have started to co-supervise more and more students that are not within our main biological system (type 2 diabetes), but working with other biological systems (e.g., liver functionality, yeast transport, heart electrophysiology, heart adronereceptor signalling, etc), and where we have only supervised the modelling side. In other words, these students have had us (usually me, Gunnar, plus one of our Ph.D. students) as modelling supervisors, and another group and supervisor for the biological aspects. Since this is now starting to grow beyond our internal capacity, and also starting to make our group a little bit dis-proportionate, we feel that the time is now ripe to externalize these activities to a separate location.

The idea is that this location should be a modelling centre, where people from different biological groups sit one or several days with their modelling, to share and discuss modelling-oriented aspects, and then sit the rest of the time in their biological group, where they can discuss the biological aspects of their work.

We will start to form this external centre at IMT, i.e., the department of medical technology. This department is situated within the university hospital, and within less than a few minutes walk from almost all biological groups at the campus. Nevertheless, they are still part of the technical faculty, and have already built up a good computer infrastructure, which has turned out to be a challenge for us to build up within the medical faculty.

We will probably start this move now within the next 3-4 weeks, and it will most likely be an ongoing discussion during at least the rest of the spring how to implement this modelling centre in a good way in practice. We will, e.g., have one such discussion meeting now this Monday, February 28, 2011, at 16-17, to which you are welcome to attend if you want.

The clear intent is also to have this as a strategic first step towards applying for major funding – both from within the university and from elsewhere – where this core-facility is part of the functionality provided by the Linköping Centre for Systems Biology. I therefore also attach some thoughts that I once wrote down on how this could be structured.

All in all, this is another important step forward towards establishing not only our group, but Linköping in general as a leading centre for systems biology. To expand our ambition beyond our own group is vital also for our own success, since many applications are network-based, and since for instance students need a bigger market than one group for their future carreers, etc.

We are now in the middle of preparing for (hopefully) four workshops to take place here in Linköping in the period of June 17-23.

One of the reasons for these workshops is that we want to make use of our newly recruited guest professors Hiroaki Kitano and Jens Timmer, by having some of the most important international players come here and discuss some vital issues with us. Kitano is one of the world-leading players in model construction (e.g. by being the founder of CellDesigner, the most widely used software in the field), and he has also recently moved into hierarchical modelling, which is the topic of one of the workshops. Similarly, Jens Timmer heads one of the leading groups of sound data-analysis approaches to systems biology modelling, including the handling of uncertainty in data; this is the topic of another workshop. The third workshop comes from a network that we are forming on multi-level hierarchical modelling of type 2 diabetes, and the final workshop is meant for us in the Linköping Centre for Systems Biology.

More information on all these workshops can be found here

So, welcome to Linköping in June! 🙂

Education-wise, we in ISBgroup have one major highlight during the year: our course TSRT17, the systems biology project course.

This course runs for third years students at TB (engineering biology), and this year we are also glad to have with us students from the I-Bio programme (biology with a more engineering and management training), and also some additional undergraduate and Ph.D. students who only take the initial part of the course for general training.

The initial part of the course consists of 5x2h lectures, 2 exercises, and a 12h computer lab. This gives them a crash course in systems biology, where they learn to put their various pieces of knowledge together into the art of model-construction and analysis.

The second part of the course is the major part of the course. Then they are divided in groups of 6-7 students, and given a scientific questions. This question is a real unanswered but interesting biological question, and is different for each group. The different projects all have the same character, however, since they are all structured around the analysis of a given set of data, where the task is to analyse this data for possible mechanistic explanations to the cause of the observations, and usually centered around two or more competing and plausible mechanistic hypotheses. The task is therefore to see which of these hypotheses that can be rejected, and to identify interesting and well-characterized core-predictions, which allows future experiments to test the conclusions, or differentiate between previously non-differentiable explanations, etc. All-in-all, the students are really doing real scientific research problems, which is really cool actually, both for us and for them.

The course will last for the entire semester, and there is a big presentation day on May 18, which is open for all, and which includes both a poster session, presentations by the students and by other invited lecturers, and by a prize-ceremony for the group with the best name and logga…! 🙂

So, welcome on May 18, to come and see the results of this years TSRT17-edition.

Now both contracts are signed, and all is finished – Hiroaki Kitano and Jens Timmer will be guest professors in our department during the next 2 years!

Hiroaki Kitano is one of the founders of systems biology, by e.g. having written some of the early hallmark reviews/commentaries in Nature and Science establishing the concept, and by arranging the first edition of the the biggest conference in the field: International Conference on Systems Biology (ICSB). Kitano is also chairman of the International Society for Systems Biology, which for instance oversees the ICSB, and we will try to make use of this fact to bring this conference to Linköping, perhaps even as early as 2013 or 2014. We have also already applied for additional Japan-Sweden funding, to – apart from the money coming with the professorship – further improve our possibilities for joint activities, and exchange of students, personnel, etc. The main topic of interaction will probably be hierarchical modelling, i.e., such models that can be developed in a distributed fashion, where different sub-models are put together in an object-oriented fashion.

Jens Timmer leads one of the biggest groups around in systems biology, with over more than 35 employees and encompassing collaborations with many different and leading biological groups. Jens has also been the Speaker for HepatoSys, the biggest national network around in Systems Biology. The most important reason why I am very thrilled about this collaboration, however, is that Jens’ group is doing their modelling and model-based analysis in a fashion that is very similar to ours. They also take a data-analysis point-of-view, where the intent is to extract as much information as possible from the data, e.g., by considering also the quantitative aspects of the data features. Similarly, they are also keenly aware of many of the most important pitfalls in such an analysis – such as neglection of the over-parametrization issue, or blind usage of analytical formulas for the statistical tests. We are also already collaborating on the development of improved methods to handle this issues (for instance through my previous additional affiliation as Senior External Fellow in Freiburg, and through our shared employement of Rikard Johansson), such as bootstraping-based statistical tests, and profile-likelihood-based confidence interval calculations.

These recruitments are the result of our headmaster (rektor) having an initiative of attracting some really high-profile people to our university, and of me having been a post-doc in both these groups, and therefore knows them personally. In short, we are very thrilled about this development, and will now start to think about various ways in exploiting this new fact as much as we can: when marketing our systems biology centre here in Linköping internationally and locally, when planning for future recruitments and exchanges of staffs and students, when designing new research projects, etc. All of these aspects now suddenly have drastically improved potential!

The results from this year’s evaluation and financing round from the Swedish Research Council (VR for Vetenskapsrådet) has now arrived (nov 2010), and it turned out that both me and Peter came out on a very good side of things.

* Peter Strålfors got an increase from already high 900 000 SEK/year to 1 000 000/year, during the next 3 years

* I got a research fellowship position (Forskarassistent) which means salary for me and additional money, in total approx 5.5 MSEK during the next 4 years.

There are several reasons why this is very good news. First, VR has the only really ambitious evaluation process in Sweden, where they really evaluate you by the science. This means that this is the best such quality stamp that you can get on the national level, and we both got almost as high as you can get (our group got ranked as “worldleading”, for instance). For this reason, it means that this increased money we got now will be multiplied by making it even easier to get money from other funding sources in the future: from the university there is an automatic co-funding of approximately the same order of magnitude; many other research agencies simply look at what you got from VR and scale their financing by that; and this outcome means that Peter will stay among the league of prioritized professors at University. Finally, my specific position means that I have a 4 year position ensured, which is a high-status position, and which generally goes beyond the post doc level. Since this is the “magic” step in scientific carreer (e.g., this year only 8% approval) this means that I now have very good chances of implementing my various scientific plans, for instance by keep building up my research group here in Linköping on a long-term basis etc.

So now we just need to figure out how to best celebrate and make use of these new possibilities! 🙂

We have just published a new
paper in JBC!

Mass and information feedbacks through receptor
endocytosis govern insulin signaling as revealed using a parameter-free
modeling framework, Cecilia Brännmark, Robert Palmér, S. Torkel Glad,
Gunnar Cedersund, and Peter Strålfors, J Biol Chem, 2010, doi:
10.1074/jbc.M110.106849

This it the first really
extensive systems biology paper coming out of our lab, and it is the
result of several years of iterations between experiments and
mathematical modelling. It builds on the same directions our previous PLoS
Comp Biol paper, which seeks to characterise the early events of
insulin signalling, i.e. the activation of the insulin receptor, and the
insulin receptor substrate-1 during the first few minutes. This
sub-system is beneficial to study because many relevant hypothesis can
be put forth which only describe this little isolated systems; in other
words, the tested models are small and strong conclusions can be made.

This
is the also the first paper where we really demonstrate our idea of
core predictions. Such predictions are defined as uniquely identified
model properties (based on the model structure and the available
experimental data), even though the individual parameter values might
not be uniquely identifiable. We use an especially developed
optimization algorithm to determine such properties, and then use these
properties to identify relevant new experiments.

The
future plans are to extend this model in several directions. Primarily,
we will be looking more downstream, to include other target proteins and
cellular processes. This inclusion has the ultimate goal to result in a
computable adipocyte and adipose tissue module, which both is
consistent with our detailed understanding of the cellular and local
data, and with the whole-body perspective, relevant for type 2 diabetes
and clinical trials. We will, however, also keep studying these early
processes more; the present results put receptor endocytosis at the
heart of the matter, and we hope to achieve further insights on the role
and relevance of this sub-process for insulin signalling, by combining
advanced single-molecule microscopy with model-based data analysis using
stochastic modelling.

So look forward to that! 🙂

We have just published a new paper, which is now available online here, and the citation details are as follows:

Zooming of states and parameters using a lumping approach including
back-translation, Mikael Sunnaker, Henning Schmidt , Mats Jirstrand and Gunnar Cedersund BMC Systems Biology 2010,
4:28, doi:10.1186/1752-0509-4-28

It is the first of a series of papers coming out from us concerning zooming, i.e., the ability to switch between two versions of the same model – one large and one small – where both models describe the same system, and where there is a mapping between the states and parameters in the two different models.

Such zooming may be useful in many situations. One example is in object-oriented modelling, where differently detailed models can easily be replaced for each other as one switches focus of interest. Another important application is when the reduced model is identifiable, i.e. where all its predictions are uniquely identified from the given data. In this case the zooming allows for the addition of interpretation to the reduced entities (states or parameters), and this is the ultimate step in the creation of a core-box model.

There will be several follow-up papers on this paper. In one of these, we extend the results to a certain type of non-linear systems. In another paper, we concern ourselves with the mapping between two general models, i.e., where the smaller model has not been obtained using model reduction. And in a third paper, we will show how the creation of a zoomable, hierarchical, whole-body model for glucose homeostasis can reveal new insights and improvements regarding our cellular experimental system for studying insulin resistance and type 2 diabetes.

So look forward to this!