AI solution for in silico small molecule discovery
Discover our next generation AI solution for discovery of hits and bioactivity
The goal of hit discovery is to identify a chemical compound that is capable of modulating a drug target. Zepto.Hit, Kantify's AI solution for hit discovery, is a next generation AI solution for hit discovery, which works for any disease target we have tested so far.
Challenges of existing hit discovery approaches
Traditional small molecule hit discovery approaches often rely on physics-based modeling or machine-learning algorithms like QSAR. Physics-based approaches can be precise but come with limitations, such as the need for many parameters to be set for good results and computational intensive nature. On the other hand, machine-learning approaches are fast but suffer from poor generalization to new molecules and have limited performance on novel targets or targets with limited known binders.
Kantify’s small molecule prediction solution Zepto.Hit
Zepto.Hit is part of Zeptomics, Kantify Drug Discovery platform. It stems from a combination of extensive Machine Learning innovations with a very large dataset. Thanks to this combination, it achieves superior predictive performance compared to both physics-based and traditional machine-learning approaches.
Zepto.Hit can quickly and efficiently recommend promising hits, without limitations to chemical structures similar to known hits. Furthermore, Zepto.Hit can accurately make predictions for proteins that have no known modulator or binder. Zepto.Hit can be seen as a very smart filter which will predict promising compounds across billions of possibilities. These compounds can later be screened and validated in a wet lab in a low throughput screening campaign.
Why is it unique?
For any protein of interest, Zepto.Hit can generate a list of likely hits, with a high accuracy.
Thanks to this wide knowledge it is capable of:
Screening billions of compounds across many families, to identify promising ones
Identifying not only on-target hits, but also off-target hits. Thanks to this, the risk of adverse effects can better be identified.
Identifying hits even in the case of targets for which no ligands are known. Also here, pharma researchers can pre-screen a novel target even if it is not well known or studied.
Prioritizing compounds that are synthesizable and available so they can be rapidly tested and validated.
Zepto.Hit is a truly groundbreaking solution, which can accelerate the discovery of drugs for any target of interest.
We validated our approach on various targets across several indications. For some, no ligand had previously been identified in the literature.
Across all projects where we have validated Zepto.Hit in a wet lab, our AI model averages hit rates between 20% to 67% even when adding strict constraints to the molecules screened (e.g. excluding molecules from the same chemical family or applying PK/PD quality criteria).
Due to its performance and capability to screen billions of compounds, Zepto.Hit is a groundbreaking model. Its benefits are numerous. First of all, Zepo.Hit is an accelerator of drug discovery for any existing target of interest. It has excellent performance in rare diseases, which is a field where we have drug discovery projects (check our pipeline).
For that reason, it can considerably decrease time and cost of drug discovery efforts, Increase chances of success through the identification of possible adverse effects due to off target hits, and decrease risk by having the widest possible search of promising leads, instead of missing out on some valuable compounds.
Contact us via the below form to learn more about using our hit prediction solution Zepto.Hit