PerMedCoE Building Blocks' Catalogue

This repository contains the Building Blocks developed within the HPC/Exascale Centre of Excellence in Personalised Medicine (PerMedCoE).

Table of Contents

• PerMedCoE Building Blocks' Catalogue
  • Table of Contents
  • List of Building Blocks
  • License
  • Contact

List of Building Blocks

1. COBREXA

   This package provides the COBREXA Flux Variability Analysis (FVA) Building Block (BB). Use this BB to analyze the viable feasibility and optimality ranges of your metabolic models.

2. Carnival

   This building block uses the accelerated CARNIVAL simulator with OpenMP and ACO solver with MPI which can be used in a HPC system.

3. CarnivalPy

   This building block uses the normal version of CARNIVAL for Python with support for different open-source and commercial MPI solvers.

4. Carnival_feature_merger

   This building block merges the results of CARNIVAL.

5. Carnival_gex_preprocess

   This building block downloads GEX data from GDSC and applies minimal transformations required for other building blocks.

6. CellNOpt

   This building block uses the accelerated CellNopt simulator with OpenMP and the ACO solver.

7. cll_combine_models

   This building block combines patient or group-specific results from MaBoSS, assessing whether the obtained profiles are appropriately clustered and can serve as predictors of disease subtype.

8. cll_network_inference

   This building block involves network inference with CARNIVAL, leveraging Omnipath, as well as DecoupleR and PROGENY results as constraints within the linear programming problem.

9. cll_personalize_boolean_models

   This building block is responsible for building patient-specific boolean models by employing the PROFILE tool and input RNA-Seq data.

10. cll_prepare_data

    This building block involves an in-house script for the primary analysis of the input RNA-Seq data, focusing on tasks such as differential expression analysis and batch effect correction.

11. cll_run_boolean_model

    This building block evaluates a single patient or group-specific model using MaBoSS.

12. cll_tf_activities

    This building block entails the inference of transcription factor (TF) activities using DecoupleR and the quantification of molecular pathways through PROGENY.

13. MaBoSS

    This building block uses MaBoSS to screen all the possible knock outs of a given Boolean model.

    • There is also a GPU enabled version of this building block using MaBoSSG for sensitivity analysis (MaBoSSG).

14. PhysiBoSS

    This building block performs a multiscale simulation of a population of cells using PhysiBoSS.

15. PhysiBoSS_invasion

    This building block is used to perform a multiscale simulation of a population of cells using PhysiBoSS.

16. Build Model From Species

    This building block generates a boolean model of interest from a list of genes.

17. Export Solver HDF5

    This building block exports the CSV files defining the problem of CARNIVAL/CellNopt to a R format file required for the simulators.

18. invasion_analysis

    This building block extracts quantifications about type of invasion from a PhysiBoSS result.

19. Meta-analysis of PhysiBoSS Output

    This building block performs a meta-analysis to determine which parameters of interest in PhysiBoSS are distributed differently among the patient subgroups.

20. ML Jax Drug Prediction

    This building block implementing the ML strategy for prediction of drug responses on cell lines, accelerated with JAX.

21. Omnipath

    This building block downloads the latest prior knowledge network (PKN) from the whole database from a predefined list of genes and can be reduced to a subset of selected public databases. It can be extended to provide more options to pre-filter genes for example.

22. Personalize Patient

    This building block tailors a given MaBoSS Boolean model to a given RNAseq dataset of interest, for instance from the "Single-cell Processing" building block.

23. Print Drug Results

    This building block generates report from raw simulation results of large drug screening.

24. Progeny

    This building block computes a matrix of samples x pathways with pathway activities given gene expression data. This is required to compute cell features summarising gene expressions into a vector of 11 pathway activities that is useful to predict drug responses for any cell.

25. Single-cell Processing

    This building block performs the processing and analysis of the single-cell RNA-Seq data from the patients included in the sample.

26. TF Enrichment

    This building block uses the tool Dorothea to calculate the TF activities from changes in downstream gene targets. This is required by CARNIVAL in order to extract a subnetwork from the PKN connecting perturbations to TFs.

License

Apache 2.0

Contact

https://permedcoe.eu/contact/

This software has been developed for the PerMedCoE project, funded by the European Commission (EU H2020 951773).