ADMET prediction with Artificial Intelligence : Publication

Many therapies fail along the clinical stage. One of the major reasons for failing are non-anticipated adverse effects.

Using Machine Learning, we can now predict with a high level of reliability what will be the possible ADMET (absorption, distribution, metabolism, and excretion) of a drug to select those drugs that have the highest chance of succeeding.

The peer-reviewed journal Frontiers in Pharmacology published on 27 April 2022 the joint research article of Kantify, I-Stem and Genethon. It demonstrates how AI can be used to decrease the risks of failure in the drug discovery stage.

Project overview

In this project, Kantify has helped accelerate the discovery of two promising compounds for a specific type of myopathy called Limb Girdle Muscular Distrophy.

The project started from the collaboration of Genethon and I-Stem. It was based on the original idea that Limb Girdle Muscular Distrophy could be treated by limiting the degradation of misfolded forms of the alpha sarcoglycan protein.

To challenge this idea, the two institutions developed a bioassay allowing to assess in high throughput the therapeutic impact of thousands of drug candidates. They had a series of interesting hits, that needed to be scored according to their potential efficacity and risk.

Minimizing the risk associated to the discovery of a new therapy

Using its ADMET prediction technology, Kantify has ranked promising hits in function of their activity/ efficacity/toxicity score.

This has enabled I-Stem to pick-up the two compounds that have a higher chance to success at a clinical stage.

Kantify’s AI technology allowed us to select very rapidly the most promising candidate. This is really exciting to see the power of this technology

Dr Xavier NISSAN, leading scientist of the study

Partners

• CECS, I-Stem, Corbeil-Essonne, France

• INSERM U861, I-Stem, Corbeil-Essonne, France

• UEVE U861, I-Stem, Corbeil-Essonne, France

• INTEGRARE, Genethon, Inserm, Univ Evry, Université Paris-Saclay, Evry, France

• Université Paris Est Creteil, INSERM, IMRB, Créteil, France

• KANTIFY, Bruxelles, Belgium

• Département de Genetique, DMU Biologie-Pathologie, GH Mondor-A. Chenevier, AP-HP, Creteil, France

• Institut Curie, PSL Research University, CNRS, UMR144, Structure and Membrane Compartments, Paris, France

• Institut Curie, PSL Research University, CNRS, UMR144, Cell and Tissue Imaging Facility (PICT-IBiSA), Paris, France

Learn more

You can access the publication on the Frontiers in Pharmacology website.

Work with us

Are you a drug hunter willing to maximize the chances of success of its drug discovery project?

Get in touch through the form below!