Descriptions

Modelling biological systems: towards digital twins
Prof. Ilja Arts
Digital version of ourselves are no longer science fiction. At this moment, scientists all over the world are building computer models that can be used to simulate the effects of treatment of an individual patient, or to better understand how biological systems like our brain work. Systems biology is a combination of biology and mathematics that tackles complex problems in a holistic way. I will introduce the concept of systems biology and give some examples of its manifold applications.

Modeling the human brain to study hearing
Dr. Michelle Moerel
How do we follow a conversation in a noisy environment? How do we recognize speech irrespective of the speaker’s accent? The human brain, arguably the most complex biological system there is, effortlessly achieves these tasks. I will discuss how brain imaging combined with computational modeling can shed light on human hearing.

Systems Biology vs. zombie cells: Tackling senescent cells in human aging
Dr. Marian Breuer
Cells that are aged, damaged or in danger of turning to cancer cells can become "senescent". In this state they cannot divide anymore but can flood their surroundings with harmful secretions. These "zombie" cells accumulate more and more as a person ages, can damage surrounding cells and can even make them senescent too. How can systems biology help tackling these cellular zombies and their role in human aging and age-related diseases?

Descriptions

Predicting the effect of single-point missense variants in the binding site on protein-ligand binding affinities: A machine learning approach
A. Ammar
A key concept in drug design is how natural variants, especially the ones occurring in the binding sites of drug targets, affect the inter-individual drug responses and efficacies by altering binding affinities. These effects have been reported in literature on very limited and small datasets that are not suitable for machine learning prediction models. Ideally, a large dataset of binding affinity changes due to binding site single-nucleotide polymorphisms (SNPs) is needed to build a machine learning (ML) model predicting these effects.

However, to the best of our knowledge, such a dataset did not exist yet. To solve this, a reference dataset of ligands binding affinities to proteins with all their reported binding site variants was constructed using a molecular docking approach. A numerical vector representation of protein, binding pocket, mutation, and ligand information was encoded using a total of 256 extracted features to describe the protein-ligand pair. Using this dataset, two designs of machine learning regression models were trained and evaluated on the chosen features to predict the binding affinity, and six different scenarios to split training and test data were evaluated. The models trained on datasets based on ligand molecular weight split reported the best performance. The best performance model reported an RMSE value of 0.57 kcal/mol-1 on an independent test set with an R-squared value of 0.86 and a prediction speed of 29 bound compounds affinities per second. We report an improvement in the protein-ligand binding affinity prediction performance of the ML models compared to several published models. The obtained models can be used in early-stage drug discovery to rapidly and accurately obtain a better overview of the ligand binding affinity variability across genetic variants. This may benefit applicability of medicine across ethnic groups as well as in personalized medicine.

The Visual Cortex as a Network of Phase-Oscillators
K.S. Evers
The human brain is capable of extracting information from visual scenes. Mathematical models of the visual system can be used to link biological structure to function. Neurons in the early visual cortex are sensitive to oriented stimuli, and local synaptic connections are biased towards neurons with collinear receptive fields. This suggests a role for lateral connections in the early visual cortex in contour integration. Human psychophysical experiments (Field et al., 1993) suggest that a local association field is expressed by the visual system. This local association field suggests a relatively high association between collinear and proximal oriented stimuli.

The higher association between collinear stimuli allows for grouping of smooth contours. In this project the local association field is implemented in a phase-oscillator network model assuming binding-by-synchrony. The phase-oscillators represent neuron populations of the early visual cortex, the lateral connections between the oscillators are based on the local association field theory, and grouping of contours is assumed to take place through synchrony between phase-oscillators. The frequency- and phase code can be used by the brain to interpret a visual scene. Experiments are performed by simulating the phase update in the Kuramoto model. The first experiment confirms that the model is capable of performing contour integration through synchrony. New experiments are designed to generate testable predictions and hypotheses. These predictions, hypotheses and future experiments and extensions of the model are discussed.

Assessing Outcomes and Risk Factors Of Atrial Fibrillation Through A Lifetime Population-Level Markov Model
Cristian Alberto Barrios Espinosa
Atrial fibrillation (AF) is characterized by abnormal electrical and mechanical activity of the atria and presents a major global health burden. Computational models are increasingly being used to study AF and guide its treatment. However, current computational models and epidemiological studies are unable to bridge the gap of knowledge between mechanisms of AF pathophysiology and long-term clinical outcomes.

Aim: To develop a novel computational model to bridge this knowledge gap by simulating a virtual population of patients over a lifetime period. Methods: A Markov model with 5 states for combinations of AF, sleep-disordered breathing (SDB) and death was developed. The stochastic transitions between states are controlled by clinical risk factors (age, sex, SDB) and mechanisms of AF pathophysiology (shortening of effective refractory period and slowing of conduction velocity). Results: The model was calibrated to reproduce real-world data with respect to: 1) epidemiology of AF and SDB, 2) mortality in the general population and in AF patients, 3) patterns of AF paroxysms, and 4) relation between clinical subtypes of AF and atrial fibrosis. Finally, the calibrated model was used to analyze the potential therapeutic effects on AF of different types of risk-factor management for SDB. Conclusion: We successfully developed a model that connects mechanistic and epidemiological research in AF and show how this model can be used to perform a ‘virtual clinical trial’ to generate new hypotheses that can be used to guide new clinical research and ultimately lead to better AF management. Impact: This novel patient-level model will be used to study the effect of new policies on the burden caused by AF.

Towards understanding complex behaviour in goal-oriented systems
PR Stolpe
Understanding complex, goal-oriented behaviour remains an ongoing challenge in neuroscience. Meaningful progress has been made through intelligently designed experiments. Visual and sensory information processing as well as action generation were found to be a function of the frontoparietal network spanning the visual, sensory and motor cortices. Computational neuroscientists successfully devised models to explain neural phenomena encountered in the frontoparietal network. However, unlike other natural sciences computational neuroscience must not only account for neural phenomena, but also address information processing in the brain. Therefore, an ideal model must be capable of generating complex, goal-oriented behaviour while also accounting for neural phenomena. Dynamical systems have been widely applied to model neural dynamics, but also to generate control policies in robotics. This work proposes to leverage computing with dynamical systems to understand neural dynamics and complex behaviour expressed as control policies in goal-oriented systems.

To that end, this work utilises recent advances in deep reinforcement learning. Agents are trained to solve a variety of reinforcement learning problems including control of an anthropomorphic hand. Afterwards, dynamics recorded while the agent performed the task are critically assessed. Attractor networks and central pattern generators are found as general computational strategies to solve for final state goals and periodic tasks (e.g. walking), respectively. These findings demonstrate the capability of computing with dynamical systems to account for neural dynamics and information processing. Computing with dynamical systems when applied to more complex tasks could pave the way to generating ever more sophisticated hypotheses about complex behaviour in artificial and biological systems.

Translational Neuromodeling, Computational Psychiatry and Computational Psychosomatics
Klaas Enno Stephan | Translational Neuromodeling Unit (TNU), Institute for Biomedical Engineering, University of Zurich & ETH Zurich
For many brain diseases, particularly in psychiatry, we lack clinical tests for differential diagnosis and cannot predict optimal treatment for individual patients. This presentation outlines a translational neuromodeling framework for inferring subject-specific mechanisms of brain disease from non-invasive measures of behaviour and neuronal activity. Guided by clinical theories of maladaptive cognition and aberrant brain-body interactions, generative models can be developed that have potential as “computational assays”.

Evaluating the clinical utility of these assays requires prospective patient studies that address concrete clinical problems, such as treatment response prediction. If successful, computational assays may help provide a formal basis for differential diagnosis and treatment predictions in individual patients and, ultimately, facilitate the construction of mechanistically interpretable disease classifications.

Predicting neuronal response properties from hemodynamic responses in the auditory cortex
Isma Zulfiqar | PhD alumnus MaCSBio, Researcher, Maastricht University

The differences in neuronal response properties between the rostral and caudal streams in the auditory cortex are thought to support the specialized functions of ‘what’ and ‘where’ (or ‘how’) processing, respectively. While the responses in rostral and caudal auditory belt regions have also been examined in the human auditory cortex using fMRI, it has been challenging to relate observed differences in fMRI responses along the rostral-caudal axis of the human temporal lobe to fundamental neuronal response properties.

To bridge this gap, we presented a forward model combination of neuronal model of the auditory cortex with physiological model of hemodynamic BOLD response. Our simulations showed that the hemodynamic BOLD responses of the caudal belt regions in the human auditory cortex were best explained by modeling faster temporal dynamics and broader spectral tuning of neuronal populations, while rostral belt regions were best explained through fine spectral tuning combined with slower temporal dynamics. These results suggest two parallel streams of complimentary information processing in human auditory cortex.

The Behavioral Arnold Tongue and the Perception and Learning of Figure-ground Distinction
Maryam Karimian | PhD alumnus MaCSBio, Researcher, Maastricht University​

We combined psychophysical experimentation with Kuramoto phase oscillator modeling approaches to investigate the role of neural gamma oscillations in figure-ground distinction in visual images. Thus, a set of stimuli constituted from Gabor annuli is presented to a group of human subjects in the experiments and also, used as the input for the computational model. Our main hypotheses are as follows:

1- The likelihood of figure-ground distinction would be a function of contrast variance and spacing scale between the Gabor annuli in the stimuli. That is investigated by constructing the behavioral Arnold tongue.

2- Through learning, the likelihood of figure-ground distinction would increase. In other words, the region in the behavioral Arnold tongue that indicates highly probable figure-ground distinction would grow.

3- Perceptual learning is specific for the position of the figure in the visual field. That means our findings would support the idea of bottom-up learning, which involves plastic changes in lower-level cortical areas such as V1.

Workshop 1: Beginner’s guide to metabolic networks: From modelling to visualization
Chaitra Sarathy | PhD candidate, MaCSBio

• Target audience: students/computational biologists (anyone curious about metabolic modelling)
• Pre-requisites: Knowledge of MATLAB and Cytoscape will be useful but not mandatory as this is an introductory workshop
• Software installation: Find the instructions here

This workshop will introduce the audience to the field of constrained-based modelling. The goal is to teach how metabolic models can be simulated and visualized. In the first part, the basic concepts of constraint-based modelling will be outlined. This will be followed by a hands-on demonstration of the tool, EFMviz, for visualizing and analyzing large metabolic networks.

Workshop 2: Network analysis in biomedical research using Cytoscape
Dr. Martina Kutmon | Assistant Professor, MaCSBio

• Target audience: biomedical researchers, students, bioinformaticians
• Software installation: Find the instructions here

Cytoscape is a widely adopted network analysis and visualization toolbox used in biomedical data analysis. In this 2-hour workshop, you will get a short introduction into the field of network science and how networks can be used in biomedical research. After a short demo of Cytoscape, you will gain hands-on experience by walking through a set of short exercises to get familiar with the tool (network creation, network extension, data visualization, layouts). For users experienced with coding in R or Python, I will provide optional automation exercises.

Workshop 3: Using deep neural network models to unravel neural sensory processing: Sound encoding in the human brain
Dr. Kiki van der Heijden | Postdoc, Radboud University, Nijmegen

How can deep neural network (DNN) models help neuroscientists understand brain functioning? This workshop gives a short introduction to the use of DNN models for investigating neural sensory encoding, i.e. the transformation from a stimulus in the world into a neural response in the brain. We will discuss the concept of DNN models and highlight several popular DNN model classes and their potential and limitations. Further, illustrated by examples I will describe the recent expansion of DNN models from visual neuroscience into other sensory domains such as auditory neuroscience. To conclude, we will examine a current research project from our group on sound location processing in the brain as a hands-on example of DNN model development in the context of neuroscience.

• Prior knowledge requirement: None, this is an introduction level-workshop aimed at a broad audience that does not require mathematical or neuroscientific background knowledge.

Workshop 4: Pathway Drawing and Curation Workshop
Friederike Ehrhart and Lauren Dupuis | BiGCaT Department of Bioinformatics

• Target audience: Wet lab and dry lab scientists interested in drawing biological pathways
• Software required: PathVisio, Java8
• Data required: BridgeDb mapping files

There is a multitude of information about molecular interactions and biological pathways in literature that can help us better understand biological processes. However, a large amount of the information can’t be utilized in data analysis - especially omics data analysis - because it is not presented in a machine-readable format. Hence, proper drawing and integration of biomedical knowledge in machine readable biological pathway models is essential for the use of the pathways by the broader scientific community. This is especially necessary when it comes to rare genetic diseases for which little information is available, and with these methods the most information and conclusions can be extracted. In this workshop organized by the European Joint Project on Rare Diseases and the Helis Academy, we will guide you through the process of human and machine readable pathway model drawing and curation with rare disease pathways. We will introduce the pathway drawing and analysis program PathVisio, walk through the process of extracting relevant information, finding appropriate ontologies, gene, metabolite and interaction identifiers, working with an identifier mapping database (BridgeDb), drawing a biological pathway, introduce the pathway repository WikiPathways, guide participants through the process of uploading and updating a pathway on WikiPathways, and teach the basics of curation. While we focus on rare diseases in the workshop, you are welcome to bring any gene or pathway of interest you wish to draw.

Towards Data-Integrated Medicine
Prof. Natasa Pržulj | Professor of Biomedical Data Science at Computer Science, University College London, and ICREA Research Professor at Life Sciences Department, Barcelona Supercomputing Center

We are faced with a flood of molecular and clinical data. We are measuring interactions between various bio-molecules in a cell that form large, complex systems. Patient omics datasets are also increasingly becoming available. These systems-level data provide heterogeneous, but complementary information about cells, tissues and diseases. The challenge is how to mine them collectively to answer fundamental biological and medical questions.  This is nontrivial, because of computational intractability of many underlying problems, necessitating the development of heuristic methods for finding approximate solutions.

We develop methods for extracting new biomedical knowledge from the wiring patterns of systems-level, heterogeneous, networked biomedical data.  Our methods uncover the patterns in molecular networks and in the multi-scale network organization indicative of biological function, translating the information hidden in the topology into domain-specific knowledge.  We introduce a versatile data fusion (integration) framework to address key challenges in precision medicine: better stratification of patients, prediction of driver genes in cancer, and re-purposing of approved drugs to particular patients and patient groups, including Covid-19 patients. Our new methods stem from novel network science approaches coupled with graph-regularized non-negative matrix tri-factorization, a machine learning technique for dimensionality reduction and co-clustering of heterogeneous datasets. We utilize our new framework to develop methodologies for performing other related tasks, including disease re-classification from modern, heterogeneous molecular level data, inferring new Gene Ontology relationships, aligning multiple molecular networks, and uncovering new cancer mechanisms.