PerMedCoE has prioritised use cases on drug synergies and COVID-19

To test its core applications and help prepare for Exascale computing, PerMedCoE has designed five biologically relevant use cases to serve as pilot projects. Among these, the ones concerning the study of drug synergies and the study of COVID-19 have been prioritised.

PerMedCoE is a project that aims at laying the grounds to solve personalised medicine problems by harnessing the boost of capacities that Exascale computation promises. If you are not familiar with what Exascale computing is, have in mind that the fastest supercomputer in the world since June 2020, according to the TOP500 list, Fugaku, achieves 4,22E17 floating-point operations per second or FLOPS. Exascale computing refers to computing systems capable of calculating at least 1E18 FLOPS, which is 2.7 times faster than the current fastest supercomputer. Exascale computing promises faster and more capable computers, but researchers need to start thinking about how they are going to use these supercomputers and how to adapt current software developments to them.

To help prepare this scenario, PerMedCoE has dedicated one of its tasks to establishing biologically relevant use cases to serve as pilot projects. These use cases will test the capacities and limitations of the four tools developed in PerMedCoE and test how they could be used in Exascale computers. However, the use cases are not only tests for the tools, but they have scientific value by themselves. PerMedCoE selected them to reflect a broad range of computationally-demanding real-life biomedical scenarios in which to use cell-level models. The focus of cell-level models is to study what mechanisms are perturbed in cells when disease happens or when a drug is detected. These models can study metabolism [1], signalling pathways [2] and perturbations [3], among others.

PerMedCoE has started working with two of the use cases: one on drug synergies and another on COVID infection.

Drug synergies for cancer treatment use case

The use case on drug synergies aims at finding effective drug combinations for cancer. To achieve this, we are using drug-response experiments and publicly available databases such as GDSC and personalised cell-level models. The study of drug combinations is a common topic in clinical medicine, as it is widely known that drugs have side effects, notably from the high concentrations used. Thus, it would be great to find better combinations that would allow reducing the drug doses. PerMedCoE is helping on this topic by studying what the effect of different drug combinations and concentrations is in the simulation of cell-line-specific models, which are not generic signalling models as you can find in online resources, but instead, capture specificities of a given cell-line and how drugs interact with it. Additionally, in this use case, we also plan to integrate the results obtained by BioExcel Centre of Excellence that, among other things, aims at finding and repurposing drugs for several diseases.

COVID-19 multiscale modelling of the virus and patients’ tissue use case

The focus of the other prioritised use case is to study COVID-19 disease using multi-scale models and single-cell data. Multi-scale modelling is a type of modelling that considers different time- and space-scales, notably integrating the metabolic or signalling pathways models inside agents that can capture different cells’ population behaviours [4]. Our multi-scale model considers an epithelial cell layer that is responsive to virus infection by signalling pathways, that recruits different immune cells and these interact with the cell and the viral particles. This model is a comprehensive way of organising all the knowledge available on mechanisms among these players, identify biomarkers and propose therapeutic targets. In addition, we plan to use single-cell data from patients to personalise these models. These kind of data are becoming more standard in the field and allow to have single-cell granularity, enabling researchers to study the diversity of different cells, their different maturation stages and how these affect their signalling and behaviour [5]. We want to integrate these data to help to explain the different patients’ severity observed in the clinics and find mechanisms that cause this diversity. In short, this use case wants to fill a gap in COVID-19 research: the lack of drugs available to clinicians. COVID-19 is a pandemic for which, after tremendous efforts, the world has a few vaccines that are great means for the prevention of infection. Nevertheless, we are still struggling to find proper drugs to use in clinics once the patient is infected.

Figure 1. COVID-19 infection in the normal scenario

Figure 2. COVID-19 infection with viral M protein mutated

With our multiscale model (Figure 1), we can study the mutations in the patients’ epithelial cells and in the virus and inspect their effect on the epithelial cells’ ability to kill themselves, a process known as apoptosis. Additionally, we can have heterogeneous cell populations, for instance, that 95% of the virus have a disabled M protein, preventing the release of the virus from infected cells (Figure 2) and we can compare this with the normal scenario (Figure 1).
Figures such as these can be obtained using our online website: https://nanohub.org/resources/pb4covid19

References

1. Gu, C., Kim, G. B., Kim, W. J., Kim, H. U. & Lee, S. Y. Current status and applications of genome-scale metabolic models. Genome Biol. 20, 121 (2019).
2. Abou-Jaoudé, W. et al. Logical Modeling and Dynamical Analysis of Cellular Networks. Front. Genet. 7, 94 (2016).
3. Hill, S. M. et al. Inferring causal molecular networks: empirical assessment through a community-based effort. Nat. Methods 13, 310–318 (2016).
4. Osborne, J. M., Fletcher, A. G., Pitt-Francis, J. M., Maini, P. K. & Gavaghan, D. J. Comparing individual-based approaches to modelling the self-organisation of multicellular tissues. PLOS Comput. Biol. 13, e1005387 (2017).
5. Song, Y. et al. Single cell transcriptomics: moving towards multi-omics. The Analyst 144, 3172–3189 (2019).

Author: Arnau Montagud (BSC)