Group projects

Participants in this course worked in small groups in projects designed by the trainers of the course. On this page, you can find the abstract of the projects, the general instructions that the participants in the course received and a link to a document with more detailed descriptions and instructions for each project.

• Group projects introduction
• Group projects descriptions and instructions

Group project using OmniPath+DecoupleR+CARNIVAL

Analysis of mechanism of action of kinase inhibitors on cancer cells from perturbational gene expression data.

This project aims to explore how kinase inhibitors affect gene expression and cellular pathways in cancer cells. Through differential gene expression analysis, estimation of transcription factor activities, drug ranking, data visualisation, and pathway activity assessment, the exercise seeks to hypothesise the mechanism of action of these drugs and support these hypotheses with scientific evidence. Automation tools are used to contextualise the findings and propose potential signalling interactions between drug targets and transcription factors.

Group projects using COBREXA

Cell line processing

This is a typical analysis project for profiling the cell lines using available genomic data, and comparing the predicted growth to the measured real results. This may serve as a validation of the model performance, and as a guiding prediction for the design of new treatments and medications.

Variability in metabolism

We know how to explore the “extreme” solutions of the model – flux variability analysis (FVA) gives a pretty good overview of what is possible to do with the model at the optimal production/growth. Some model perturbations such as gene knockouts do not change the optima, but have indeed an effect on the reachable states in the model. Here we use FVA to detect the variability/reachability changes caused by knocking out genes from the models.

Metabolic adjustments in bacteria

Metabolic adjustment is a phenomenon where the metabolic regulations in the bacteria tend to use the minimal possible regulation available for reaching the optimal growth. In turn, since multiple possible regulation configurations are possible for the optimal growth, the phenotype of the bacteria that is present may often depend on its growth history (e.g. propagation media) that are not available anymore. This project explores these effects. Note: this is an advanced project.

Group projects using MaBoSS and PhysiBoSS

Mapping mutants of cell cycle using cell lines data and multiscale modelling

Students will use cell line omics data to personalise a cell cycle model published by Sizek et al. 2019, and study the effects of drugs on the control of the cell cycle. First, students will use MaBoSS to study the effects on the intracellular signalling model, and find interesting single or double mutants. Then, they will use this model in a PhysiBoSS simulation to study the effect of these mutants in a population of cells.

Studying drug interactions in intra- and inter-cellular modelling

Students will use prostate cell lines omics data to personalise a prostate model and find the best drug targets for each cell line using MaBoSS intracellular modelling tool. Then they will bring those models to the PhysiBoSS agent-based modelling framework and simulate the drug effects in populations of cells. This project will reproduce part and go beyond two papers: Montagud et al. eLife (2022) and Ponce-de-León et al. bioRxiv (2023).

Studying the effects of having dynamic drug treatments using TNF

Students will use the Calzone cell fates model to investigate the best drug treatment for 3T3 fibroblasts using PhysiBoSS. They will inspect the mechanism that transmits the TNF binding with the Boolean model and perturb it. They will also explore different drug regimes by changing 3 parameters. This project will reproduce part of and go beyond Letort et al. Bioinformatics (2019) and Ponce-de-Leon, et al. Frontiers in Molecular Biosciences (2021).

Studying mutants in a multiscale model of lung epithelium infected by COVID-19

Students will use single cell data from bronchoalveolar lavage fluid to identify epithelial and macrophage data to personalise the corresponding Boolean models. Then they will use MaBoSS to browse all possible mutants and identify the best ones to simulate them in PhysiBoSS. Lastly, students will test different virus loads in different mutants and their effect in the different models.