The ISBGroup Blog

Cardiovascular disease (CVD) is one of the most life-threatening conditions of our time and one with limited treatment options for severe cases. Improved diagnostics and preemptive treatment methods are thus especially important. Fredrik Eklund’s thesis, labelled “Guided Bayesian Graphical Network model for Cardiovascular Disease Prediction”, continues the work towards a tool aimed to provide decision support for medical practitioners.

Bayesian Graphical Networks (BGNs) are prominent decision components in the field of machine learning and artificial intelligence. A strength lies in their ability to infer partially missing data using its distribution and other data, i.e. from the same patient. This is a powerful feature in clinical practice because some informative measurements may be costly or invasive to obtain.

The thesis report suggests rules to model BGNs by in order to battle issues with high dimension clinical data as well as to reflect the pathophysiology of CVD. An algorithm was then developed to build BGNs according to these rules. Following these two steps, a BGN to predict CVD was built from clinical data on type 2 diabetes patients. The result is a network which joins prior knowledge with machine learning techniques. The network can be considered a proof of concept and a mild success.

According to agreement with Gunnar Cedersund, Fredrik Eklund will continue work on the model in the months to come.

Last Wednesday (14/6) I presented my master thesis entitled Multi-level
and Multi-scale Modelling: From Adipose Intracellular Insulin Signaling to Whole Body Metabolism and
Weight Alteration. This presentation included a brief explanation of all my
work here as a part of ISBGroup. I started my journey here as an intern already
in the summer of 2015 following the bachelor course. Back then, I worked with a
project involving the development of a model describing the intracellular signalling
chain in rat adipocytes. The internship lasted for two months, but the project
went so well, that I decided to continue the work alongside my studies. One and
a half year after the start of the internship, the project successfully resulted
in a published article. Around that time, I was supposed to start writing my
master thesis, so it felt natural to continue on this track. So, by using the
newly developed intracellular model as a base, I continued the development towards
a first multi-level and multi-scale model in rodents. The abstract from my
master thesis is attached below.

Type
2 diabetes, T2D, is a prevalent, costly disease that affects millions of people
worldwide. It is a complex metabolic disorder which is characterized by insulin
resistance. The resistance is believed to initially arise from adipose
intracellular malfunctions and later spread to adjacent tissues, eventually
affecting mechanisms on multiple body levels. To handle this complexity,
mathematical models has been developed through systems biology modeling. These
include models, in both rodents and humans, that describe mechanisms on single
body levels, as well as more advanced models that describe T2D associated
mechanisms on multiple levels. However, these multi-level models have so far
mainly revolved around humans, and no such model focused around diabetes in
rodents has yet been developed. Animal models possess many benefits, like
processing valuable animal data that is difficult to obtain from human
experiments, or act as replacement for test-animals. For these reasons, a first
multi-level and multi-scale rodent model has been developed that describe
mechanisms on multiple levels; from the intracellular level (the adipose
insulin signaling chain from the insulin receptor, IR, to the GLUT4 transporter)
to the whole-body level (including the glucose-insulin metabolism and weight
change). Experimental data was obtained on these various levels, and used to
improve and test the model. The intracellular section of the model was created
through a connection of two existing models and can describe over 140
experimental data points. It was also validated with independent experimental
data and was used to form predictions about the relative contribution of the
pathways from IR to GLUT4. The new multi-level model was developed after new
connections was established between this intracellular model and a model
describing meal responses over a couple of hours, as well as with a model
describing weight change over days and weeks. The resulting multi-level model
have created insights to what mechanisms are necessary in order to simulate
data on all levels. The model also aims to increase the mechanistic
understanding of T2D and eventually provide a contribution to the drug
development process, by reducing the amount of animal trials and increasing the
success chance of the developed drugs.

– Niclas Bergqvist

This spring, we welcome a new student in our group: Thirza Poot

Introduction:

Last
spring, I finished my third year of the Engineering Biology study program at
Linköping University with a bachelor’s project within systems biology. After
this, I decided to take a break from the study program in Linköping to study in
Vienna for the winter semester 2016. Recently, I returned to Sweden, and have
decided to join ISB group for another project within system’s biology!

In the
bachelor’s project, I worked with two models regarding glucose concentrations
in critically ill patients admitted to the intensive care unit (ICU). This
semester, I will instead work with models regarding the interplay between non-esterified
free fatty acids (NEFA), glucose, and insulin, and their connection to obesity
and type 2 diabetes. The goal is to also investigate if and how these models
can be adapted to the ICU.

Earlier this month I attended the annual meeting of the European Society for Magnetic Resonance in Medicine and Biology. At the conference I had the opportunity to giva a presentation of a part of my thesis project. In this part of the project, we’re using mechanistic modelling of data from contrast enhanced magnetic resonance imaging in order to get biomarkers that can quantify liver function. The idea is that our biomarkers should be used by physicians, e.g. when planning liver surgery.

/Markus

Picture of me presenting.

During the autumn we have three new students in our group
Nicolas, Fredrik, and Johanna: The latest students to join our group.

Here’s their introduction

Nicolas

I’m
a student at Linköpings university, studying for a masters of science in
engineering biology, I have finished my third year but have decided to take a
break to work as an intern at the ISB group.

The
project I’m involved with at the moment revolves around studying the regulatory
mechanisms that control the antioxidant levels in the body. Antioxidants are
generally believed to be unregulated and simply accumulate upon continuous
intake and such an accumulation could help explain the many health benefits
observed when eating antioxidant rich foods. However this accumulation has
never been observed for more than a few hours after intake. The results of
previous studies show that there is reason to believe that the antioxidant
levels are regulated by homeostasis. The purpose of this project is to use
systems biology to investigate the regulatory mechanisms involved with
antioxidant metabolism and see if there is a homeostatic regulation of the
body’s antioxidant capacity. This project will continue and expand upon the
project I did for my bachelor thesis in the spring of 2016 where I investigated
the regulatory mechanisms that control the antioxidant capacity (AOC) in the
cytoplasm and used mathematical modelling and prediction analysis in order try
to predict how the AOC levels would change after a single dose of antioxidants
are given to a test subject. At my disposal i had access to data collected from
experimental studies conducted at Newcastle university. With the results from
my bachelors project I was able to provide the team in Newcastle with a decision
basis for how to design a new experimental study which they conducted in the
summer of 2016. I will use this data, collected from the latest study in
Newcastle to continue investigating how antioxidant capacity is regulated and
also investigate how different test subjects react differently using nonlinear
mixed-effect modelling.

Fredrik

My
name is Fredrik Eklund and my background lies in bioengineering, with a focus
on materials and biosensors. I’m currently working on my master’s degree thesis
in the field of systems biology, however. The project I’m involved in is a
continuation of previous students’ work on a statistical heart attack model. It
is a predictive model meant to be used as decision support by doctors. In its
core, a Bayesian network is trained to understand relationships between a large
number of parameters. The trained network is then used on a new patient’s
measurements for prediction. Based on the result, a medical practitioner can
then take the appropriate course of action.

Johanna
My name
is Johanna Fridberger and I am a new member of the ISB group. I will be staying
for one semester conducting a project on the model developed by Fianne Sips described
in the article ”Model-Based Quantification of the Systemic Interplay between
Glucose and Fatty Acids in the Postprandial State” in which I will exchange the current NEFA (Non-Esterified Fatty Acids)
part of the model to one that better describes physiological conditions. This
is possible thanks to new data on NEFA homeostasis. I am doing the project as
part of my education in medicine at Linköping University, now reaching the end
of the third year. I have really enjoyed working in this group so far, and I am
excited over the next couple of months here!

I have just jumped on the train from Linköping, Sweden, heading south, and this train ride marks the beginning of an intensive 4 week travelling period, entailing 4 conferences, 6 conference presentations, and numerous visits to groups all across Europe. If you want to catch and meet us along the way, read on below!

Figure 1: Just as in an earlier blog post, I am writing this blog post on the train. This time, I have an interrail ticket and some of my favorite cheeses in front of me!

My first journey heads down to London, where I will attend a workshop on September 14-15 I just very recently discovered: “Accelerating the acceptance of mathematical models as evidence in safety and efficacy decision making”. Apart from some individual exception, this will be a mostly new society of scientists I haven’t interacted with before, but the topic is one that lies close to my heart: how to incorporate mathematical models in drug development and certification. As some of you may know already, we work closely together with AstraZeneca, who also finance a position in my group, on exactly this topic. At the meeting, I am looking forward to hearing more about the Padova diabetes simulation engine (which we have used as a basis for our multilevel type 2 diabetes models), and about modelling for cardiac safety assessments. Both of these are examples where modelling as a replacement for test animals has gone a long way, and it will be nice to hear some updates, and to hear what else has happened in this field recently. This will be the only workshop where neither me nor somebody from my group presents with an oral presentation, and the reason for that is that I signed up only a few days ago, when the programme was already set.

Figure 2: This a picture from the event that in many ways can be seen as the inaugural event of my group: ICSB in Stanford 2009. Since then we have gone to almost all of the ICSB conferences.

My second journey continues on south, down to Barcelona and the International Conference of Systems Biology (ICSB), which is held September 16-20. ICSB (link) is the biggest annual conference in the field, and since it is the first time in a few years that the conference is held in Europe, it is important for us attend. The conference is arranged by the International Society for Systems Biology, headed by Hiroaki Kitano, who also was a guest professor in my group for a few years. We will go there quite a few of us from our group – me, William Lövfors, Elin Nyman, Sebastian Sten, Karin Lundengård, and Johanna Fridberger – and then there are anyway 3 people who intended to go but ended up not going (Maria Engström, Hao Li, and Rikard Johansson). At this conference, many people will be familiar faces, and we are looking forward to meeting both them and new scientists. At this conference we will also give two oral presentations:
1) Karin Lundengård with the presentation: “A mechanistic model for investigating the biological mechanisms behind fMRI“. This is based on our published model for describing the BOLD response in fMRI, and also goes into how it can be used to more correctly and precisely estimate neural activity. The presentation is held 12.15-12.35 on the Monday, in Hall 5.
2) William Lövfors with the presentation: “A phosphoproteome-wide mechanistic model of insulin signaling“. This is the updated version of the story already described in a previous blogpost, and it is held on Sunday 17.35-17.55 in the Auditorium.

Figure 3: The loggo of the Virtual Physiological Human conference.

The third conference is the Virtual Physiological Human (VPH) conference, held in Amsterdam September 26-28. This is a bi-annual conference devoted to multilevel modelling of biomedical, and often even biomechanical, systems. The VPH community is headed by the new director Adriano Henney, who also leads the Avicenna Alliance and is/was the Speaker of the german systems biology network “The Virtual Liver Network”. Before that, the VPH grew out of successful lobbying on the EU level to create a European version of the Physiome project, which e.g. pioneers the multilevel biomechanical and bioelectrical models for the heart. We will go to this conference for the second time, and our troup this time includes me, Rikard Johansson, Tilda Herrgårdh, Hao Li, and Tim Beishuizen. Originally, Elin Nyman also intended to go, but she will already have left for Boston by then. For this reason I will take her oral presentation, and thus give two lectures, along with a third presentation given by Rikard Johansson:
3) Markus Karlsson/Gunnar Cedersund “Meta-modelling combined with non-linear mixed-effects modeling for fast and robust estimation of biomarkers for diffuse liver disease”, Monday September 26, 14.20-14.40 in Emmazaal
4) Rikard Johansson “Model predictive glucose control in intensive care:
assessment in realistic clinical conditions”, Monday September 26, 15.40-16.00 in Emmazaal
5) Elin Nyman/Gunnar Cedersund, “The Virtual Adipocyte – from Big Data to
simulations of human disease”, Wednesday September 28, 09.40-10.00 in Emmazaal

Figure 4: My first slide, which I added just before holding my presentation at the last edition of ISGSB, in Durham, UK, 2014. Then I had noticed that the ISGSB has something that I really like: lots of social events, and informal activities, and a very relaxed and open athmosphere that really makes people get to know each other. The man down to the right is Stefan Schuster, who is the main arranger of this year’s edition. However, even though he was a part of the competition last year, I am not sure he has picked up on the idea for this year’s event. Nevertheless, an informal music session is on the programme, and they do have a grand piano there, which I plan to make use of!

The fourth and final conference planned to be a part of this conference is the International Study Group for Systems Biology (ISGSB), in Jena, October 4-7. There I (Gunnar) am an intricate part of the arrangements, since I am part of the ISGSB board, and since many decisions are taken jointly by this board. We are going also to this event (link) for the second time, and we will be a troup of 3 people: me, William Lövfors, and Sebastian Sten. Here Sebastian will give our final lecture for this tour:
6) Sebastian Sten, “Investigating hypotheses describing the negative brain responses in fMRI using a systems biology approach”, Friday October 7, 10.25-10.50.

Finally, and perhaps most importantly, I, Gunnar, will travel between all of these conferences by train, and will stay in Europe this entire period. In other words, I plan to visit quite a few research groups, and discuss projects and possibly also give some more lectures in local seminar series. If you are interested in meeting up with me either at one of the above events, or in between some of them, send me an email at gunnar.cedersund@liu.se . I will probably mostly be in France, England, and the Netherlands between September 20-28, and then in Germany and Austria between September 29 and October 7. But even that may be flexible to some degree!

I am looking much forward to this new travelling period, and I am looking forward to meeting many of you – colleagues and friends – at various places along the way!

Spara

Before this summer, Elin received a 2-year career support grant “International Postdoc” from the Swedish Research Council. This grant allows junior researchers to go abroad and work in world leading research environments to improve individual qualifications and extend their network of collaborators.

Elin applied for the grant together with Chris Sander at Harvard Medical School in Boston, MA, and Emek Demir at Oregon Science and Health University, OR, and will start this fall in the Sander lab. The research project is about using large-scale data handling to improve drug treatment in pancreatic cancer.

During this project, Elin expects to broaden her knowledge in large scale data handling, machine learning approaches, and cancer biology. And on top of that, being in an atmosphere like the one at Harvard Medical School, with famous researchers and the newest experimental techniques, will absolutely give her many new and exciting experiences.

Two new students are joining us this summer. Here is their introduction:

—————————————————————————————

Hello!

Our names
are Erik and Martin and we are working here at the ISB-group over the summer.
We thought that it might be appropriate to tell you about who we are, and what
we are working with!

We are both
students here at Linköping University, studying engineering in biotechnology. We
have both just finished our third year and completed our Bachelor’s project
together under the supervision of Gunnar.

During the
summer, we will continue with the bachelor’s project in cooperation with
AstraZeneca. The project revolves around creating and evaluating a mathematical
model over how the maturation of macrophages can be regulated, and throughout
this, be able to control the immune response in the immune system. If we were
able to control the immune response we could for example limit the symptoms of
auto immune diseases, in which there usually is a too high pro inflammatory
response.