Skip to main content

Translational quantitative systems toxicology to improve the understanding of the safety of medicines - Sofia: 116030

Periodic Reporting for period 4 - TransQST (Translational quantitative systems toxicology to improve the understanding of the safety of medicines - Sofia: 116030)

Reporting period: 2020-01-01 to 2020-12-31

The TransQST project gathers existing and generates new data under the project goals to support the development of tools for the assessment of drug candidates´ safety profile before undergoing clinical testing. This will be achieved by an increased understanding of when results in nonclinical testing can be reliably extrapolated to humans, and by developing models to predict exposure-based biological responses in human tissues, which is an important element of safety assessment for new drug candidates.

The philosophy underlying TransQST is that improved translation from nonclinical to human safety during clinical trials can be achieved with novel Quantitative Systems Toxicology models.

To achieve this ambitious goal, the TransQST partnership is focused on liver, kidney, cardiovascular and gastrointestinal-immune systems, common target organs for drug-induced injury to:
• Build on existing physiologically-based pharmacokinetic/pharmacodynamic (PB-PK/PD) models to define systemic as well as specific organ/cell exposure to drugs and metabolites in a holistic fashion.
• Develop SYSTEMS models for drug-induced organ damage across the four target organs.
• Integrate PB-PK/PD models and output from SYSTEMS models into quantitative systems toxicology models.
• Test the models using selected compounds with nonclinical and human data.
• Form a unique public-private partnership that leverages industry data and practical experience with public expertise in mechanistic work as well as modelling across scales of complexity.
TransQST continues to work towards better drug assessment and safety. Over 40 artefacts, including Open-Source tools and models have been developed by the consortium, and 34 articles have been published describing the project work to date. During the 4th year, the consortium members have also contributed to the fight against the COVID-19 pandemic. Some examples of our achievements are as follows:

• Development a Data Management Platform
To support long-term sustainability, one of the main aims is to provide a lightweight data infrastructure layer for TransQST. Several key tasks have been carried out, including parameter enhancements to the BioModels resource, a major dissemination platform for the open access models developed in TransQST. A public version of the BioModels resource was released and published (Glont et al., 2020) in 2020 and is publicly available at

• Development of bioinformatic tools:
- TransQST has contributed to the creation of COSMOS (Causal Oriented Search of Multi-Omic Space), an open-source tool that integrates phosphoproteomics, transcriptomics and metabolomics data. COSMOS builds on another TransQST-contributed tool, CARNIVAL, and uses optimization to integrate prior knowledge & perturbation data to reveal molecular mechanisms and pathways activated by the applied drugs. COSMOS is freely available at
- The open-source tool Omnipath has been updated & extended in year 4 by including more resources about intercellular communication and entity annotations (Cecarelli et al 2020). OmniPath is a comprehensive collection of molecular biology prior knowledge from 103 databases, with focus on literature-curated human & rodent signalling pathways. The existence of Omnipath is prior to TransQST but it has been refined & enhanced by the consortium for the benefit of all scientific community.
- A new version of the DisGeNET knowledge management platform has been released (v7.0). DisGeNET is an Elixir Recommended Interoperability Resource and contains a comprehensive collection of genes & variants associated with human diseases and phenotypes, including drug adverse events. Its existence is prior to the project, but it has been updated and improved by the consortium. DisGeNET is freely accessible at and has been published (Piñero et al. 2019). As part of the collaborative efforts towards fighting the pandemic, new section with text-mined information related to COVID-19 has been included (
- Work is ongoing on the TransQST-developed WGCNA-based TXG-MAPr web tools in collaboration with EUToxRisk and eTRANSAFE projects. Association of WGCNA module changes that are predictive of liver/kidney pathology is expected to impact lead optimization & enable prioritization of compounds less likely to induce pathology. The tools are available to the project partners ( Several industry partners have shown interest in the application of these tools for analysing their own data sets. The aim is to transfer a first version of the tools for EFPIA in-house testing in 2021, thus improving their uptake & impact.

• Development of different mechanistic, multi-scale models, that describe specific toxicity mechanisms relevant for each organ: liver, kidney, heart and gastro-intestinal system. Examples of models developed:
- The Heart WP has developed, calibrated & validated 2 new computational models: a) human ventricular cardiomyocyte electro-mechanics (Margara et al., 2020); b) human cardiac Purkinje cell electrophysiology (Trovato et al., 2020).
- A machine learning-based model for prediction of drug-induced liver injury, developed in collaboration with eTRANSAFE project, has been published (Aguirre-Plans J, et al. 2021) and is available at
- To illustrate how Quantitative Systems Pharmacology models can be employed in real-life drug safety evaluations, TransQST proarrhythmia model was used to assess antimalarials administered in the 1st wave of COVID-19 (Delaunois et al., 2021).
TransQST actions aim to generate significant impact in different dimensions by a deep understanding of the physiological, pharmacological and toxicological relevance of data and models for predicting clinical Adverse Drug Reactions (ADRs) in four target organs: Liver, Kidney, Heart and Gastrointestinal immune tract.

The project members are working towards improved methods to visualise and analyse large and complex datasets covering different types of available information (i.e. drug metabolism, transcriptomics, proteomics, metabolomics, toxicology, pathological phenotype, biomarkers) to aid decision-making on drug safety.

In this fourth year of the TransQST Consortium, different mechanistic, multi-scale models are being deployed in the organ work packages that describe specific toxicity mechanisms relevant for each organ. A BioModels resource, a major dissemination platform for the open access models developed in TransQST has also been released, and the project members have engaged in the common fight against the COVID-19 pandemic, with a drug study developed and published from the Heart workpackage.

The models cover different scales of biological organization and integrate different modelling approaches. In particular, the integration of PBPK (physiologically-based pharmacokinetic) models with QST (quantitative systems toxicology) models provides predictions of organ injury and recovery over time. These predictions are instrumental to support decision‐making, trial optimization and dose schedule guidance in early stages of drug development and it is expected that such models can be used to supplement regulatory submissions to provide additional evidence of predicted human safety response.