Modelling cell signalling pathways

Introduction

Biological systems are complex in nature and their properties often emerge from a complex interaction between their components. Therefore, predictive computational models are very useful to fully understand the behaviours of biological systems and generate hypotheses.¹ Models of biological processes, such as cell signalling, metabolic and gene regulatory networks, have been shown to divulge mechanistic insight into cellular regulation, disease formation and drug action.

In this exercise, firstly you will be introduced to BioModels,² a repository of mathematical models from published literature. You will learn to browse and explore relevant signalling models. Then, to familiarise yourself with the concept of kinetic modelling, you will start working on the cell signalling pathway, in particular, MAPK (Mitogen activated protein kinase) activation pathway (in challenge 1) which is known to regulate cell growth, survival, apoptosis, inflammation etc.^3,4

You will also learn to use COPASI,⁵ a user-friendly software to encode, simulate and annotate models to make a quantitative prediction of various scenarios to answer biologically relevant questions.

By analysing the signalling pathways, you will evaluate various scenarios such as dose dependent impact of input signal on pathway output biomarker with analysis of intermediate key players in signalling pathway, the impact of species/parameter knockdown or over-expression on pathway output etc.

Furthermore, you will get the chance to select your own model of interest and do the curation for it. Post curation, you can submit your model to BioModels database where curated models are often reused or repurposed by the broader scientific community.

Model Curation

Curating models of biological processes is an effective training in computational systems biology, where the curators gain an integrative knowledge on biological systems, modelling and bioinformatics.

Model curation requires encoding models in community standard formats such as SBML (Systems Biology Markup Language) if submitted in other formats, cross-verification of model representation and ensuring that the numerical results of the simulation correspond to the reference publication. The reproduced curation figure, together with comments from the curator on the simulation experiment and software used, is also provided to the users. For most curated models, BioModels offers the associated SED-ML file, a community standard for simulation experiment description, along with the COPASI file used for simulation. To maximise the impact and reusability of the models, the MIRIAM guidelines were proposed and the models are annotated following them in BioModels. Here is an overview of the curation process:

Recommended Pre-reading

1. Mathematical modelling:
   • https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4604653/
   • https://pubmed.ncbi.nlm.nih.gov/12648679/
   • https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3860444/

2. Software (Copasi) Tutorials:
   • http://copasi.org/Support/Video_Tutorials/
   • http://copasi.org/Support/User_Manual/
   • https://www.youtube.com/watch?v=mCPObDj-EOM

3. Antimony and Tellurium
   • Antimony
     • https://www.ncbi.nlm.nih.gov/pmc/articles/PMC2735663/
     • https://tellurium.readthedocs.io/en/latest/antimony.html
   • Tellurium
     • https://tellurium.readthedocs.io/en/latest/index.html
     • https://tellurium.readthedocs.io/en/latest/notebooks.html#

For additional/extra reading:

• https://academic.oup.com/bib/article/14/4/411/192938
• https://www.sciencedirect.com/science/article/abs/pii/S0958166900000793
• https://www.sciencedirect.com/science/article/pii/S2405471220302416#ack0010

Project aims

Curation of mathematical model of specific relevance taken from published literature/BioModels.

Steps to follow:

1. Model Selection: Select a model from Published literature or BioModels database (non-curated) as per interest and analyse the same for ODE (Ordinary Differential Equation (represented as mathematical expression in COPASI)), parameter values and initial conditions. Also look for output simulation plots and analyse them.

2. Encoding ODE using COPASI: Start encoding the model in COPASI in the form of reaction, species and parameter. Make sure mathematical expression and parameter/initial condition values are encoded correctly. Define rate laws for each reaction along with associated parameter values. If parameter values are used at multiple places, define it as global quantities.

3. Simulate encoded model: For simulating the model, you need to determine the run time, step interval and plot output specifications as per literature specifications.

4. Reproduce the literature result: Post simulation, you need to compare the literature and model plots for similarity both for x axis and y axis. Try to reproduce maximum number of plots from reference literature.

5. Model annotation: Semantic enrichment of models with cross-references to controlled vocabularies such as GO, ChEBI, Mathematical Modelling Ontology, Systems Biology Ontology, Brenda Tissue Ontology and Experimental Factor Ontology, as well as data resources such as UniProt, Ensembl, NCBI Taxonomy, Reactome etc.

6. Submit and Publish in BioModels: Post annotation, the model needs to be submitted in BioModels using user credentials.

7. Learning model development and simulation using Antimony/Tellurium (Optional) – Language and tool for encoding and simulating model for people comfortable with scripting in python.                

Challenge 1

1. Get a good understanding of basic mathematical modelling and all basic ingredients (species, reaction and their associated parameters, ODE equations) required for the creation of mathematical model of MAPK activation pathway.
2. Download the files for BioModels id: MODEL1204280001 and try to re-create the model using data in the paper and model file in COPASI. Please read the literature paper (PMID: 22748295) carefully before encoding the model. To download model files, go to https://www.ebi.ac.uk/biomodels/MODEL1204280001#Files and download .xml file. Now import the .xml in COPASI by going to file -> Import SBML option. Explore model file.
3. Double check the mathematical expression/ODE, species concentration, parameter values etc for correctness.
4. Reproduce the literature figures using your model encoded in COPASI.
5. If able to reproduce the figure, proceed to model annotation – please refer to https://drive.google.com/file/d/1JqjcH0T0UTWMuBj-scIMwsyt2z38A0vp/view?usp=sharing for annotation guidelines followed by BioModels.
6. Do a quick what-if analysis for the model (example: knockdown/over-expression of a species/parameter) and record your observations. Cross check your model and export SBML + SEDML file for the model. Also keep COPASI file. Submit all xml as main file and COPASI+SEDML file as additional files. Please refer to the link for model submission guidelines: https://www.ebi.ac.uk/biomodels/model/submission-guidelines-and-agreement
7. Prepare a presentation for the biological question answered by the model.
8. Optional Step: Encode/load model using tellurium (as package in python/Tellurium own notebook) and try to simulate model to reproduce model results.

Challenge 2

1. Select a model of your area of interest from a published literature or non-curated models from BioModels. Try to curate the same following challenge 1 steps.

References

1. https://pubmed.ncbi.nlm.nih.gov/25645874/

2. https://www.ncbi.nlm.nih.gov/pmc/articles/PMC7145643/

3. https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3063353/

4. https://www.sciencedirect.com/science/article/pii/S0167488910003228

5. http://copasi.org/

Project mentors