Project

Systems biology (SB) has proven its usefulness in projects since it can include vast amounts of data and integrate them in Boolean models for drug discoveries, high-throughput mutant analyses and patient-specific outcomes. However, simulation of biological models struggles to address large-scale simulations such as real-sized tumour simulations, the temporal evolution of microbial communities or clone-specific drug discovery. Despite some efforts, High-performance computing (HPC) uses in SB modelling are still scarce, under-optimal and not reproducible as SB modelling tools rarely use parallel processes and those efforts are usually one-shot adaptations. Likewise, Artificial Intelligence/Machine Learning (AI/ML) methods are being used to analyse the novel single-cell data with notable success, but these methods have not been widely adopted in modelling projects.

Therefore, the aim of this project is the design and implementation of a toolbox that integrates SB modelling tools that are optimised for HPC and that facilitate the use of AI/ML methods. This toolbox would scale-up the scope of projects considered in SB, allowing modelling multi-body interactions, better reuse of the vast amounts of already-published data and implementing computational standards for complex experiments in SB.

This proposal expands on my previous interdisciplinary work to provide the SB community with a flexible toolbox of advanced SB modelling software that demonstrates its usefulness in relevant biotechnological and biomedical applications. This work will bridge such fields as modelling in SB, data analyses, AI/ML methods and HPC computing, using the latter to boost the former. First, I will continue the development of HPC-optimised multiscale modelling software that allows for the integration of different modelling tools as add-ons, such as Boolean modelling, metabolic modelling, drug internalisation, etc. Second, to showcase the boost in capacities of the toolbox in SB projects, two use cases will be used as scientific benchmarks: multiscale simulations of SARS-CoV-2 infection of human alveolar tissue and real-sized bacterial communities’ interactions and protection against drugs. Finally, efforts will be dedicated to dissemination activities, which are key in the present project to involve the SB community in the HPC use.