Project Overview

The toxicology field is facing a critical need for fast, efficient and accurate technologies to assess the effects of an Increasing number of drugs and chemicals. While computational toxicology offers cost-effective and fast testing methods, It faces limitations.

Current predictive models, such as quantitative structure-activity relationship models. rely on large sets of molecular descriptors. However, this methodology can assess chemicals similar to those used in developing the models, while the large number of descriptors is hard to interpret

The EIC-funded
QUANTUM-TOX project

will develop a new type of descriptor based on quantum mechanics that can cover the entire chemical space with easily interpretable parameters. By creating specific electronic signatures and leveraging AI, QUANTUM-TOX’S innovative approach could significantly improve computational toxicology.

Project Objectives

1

to develop the first generation of ESigns targeting computational toxicology predictions.

2

to develop an AI system for toxicity prediction using the ESigns.

3

to create a software engine, codifying the lessons derived from the previous objectives to make the results available for further extended use.

4

to train and test the AI models using the ESigns for two liver conditions: steatosis and fibrosis.

Expected Outcomes and Impact

The QUANTUM-TOX project will unveil novel pathways in computational toxicology, leading to a fresh perspective on toxicological research. It promises enhanced control over the variables influencing adverse effects, signifying a significant departure from current methodologies, while being able to address the major limitations of existing approaches.

QUANTUM-TOX technology with the ability to reach the entire chemical space will help reduce the use of animals in testing, and will provide reliable means of testing environmental hazards that currently go untested

Microscope - Robot

To have enhanced
interpretability:

The use of meaningful chemical information will have a dramatic heuristic value, providing interpretability, confidence, and potentiating exploitation of the chemical knowledge. The studies will progressively move away from the empirical association between opaque parameters, entering the phase of use of explicit chemical knowledge.

To create a software engine, codifying the lessons derived from the previous steps, to make the results available for further, extended use. We will generate new models implementing the tools developed, and this will also provide a way to optimize and refine the software parameters used in the models for the toxicological endpoints.

To be able to investigate the entire chemical space

The QUANTUM-TOX technology with the novel concept of electronic signatures (ESigns) as property descriptors can investigate the entire chemical space that otherwise is not possible.

Impacts

The new system based on quantum chemistry represents a massive departure from existing approaches that depend on a limited number of compounds with reliable experimental data, being able to cover the whole chemical space. The QUANTUM-TOX project will have major scientific/technological, societal and economic impacts

Impact

To provide effective science-based safety and risk assessments that will increase and improve the screening of human and animal drugs, food products, and environmental pollutants.

Impact

It will be pivotal to accelerate the transition to innovation without the use of animals in research, regulatory testing, and education.

Impact

It will disrupt the toxicology market and enable a technology shift from benchbased toxicology to computer-based assessment.

Impact

It will decrease the attrition rates in drug discovery, enabling new drugs to market faster and affordable

Impact

It will support the new approaches in toxicology (AOPs, NAMs, etc)

Impact

It will directly affect cheminformatics through new fast algorithms that can describe large systems, thus offering a replacement for the limited methods of today.