Artificial Intelligence for Target Identification

Discover our next generation AI solution Zepto.Target, which deconvolutes therapeutic targets based on the output of phenotypic screens
Fri 10 Mar 2023

Kantify's AI solution Zepto.Target unlocks an essential bottleneck of drug discovery: target identification.

Target identification is a key step in the drug discovery and development process. The goal of target identification is to identify biological targets that play a critical role in the disease. It also aims at demonstrating that modulation of the target itself can generate the desired therapeutic effect with limited or no side effects.

Challenges of target identification

A large cause of failure during the clinical validation stage is due to the use of improper targets. What is the use of finding the right compound, if you are not modulating the right disease target?

One way researchers circumvent having to make hypotheses about the right target is to use phenotypic assays, which are assays that test the effectiveness of compounds in an in-vitro biologically relevant model of the disease, such as induced pluripotent stem cells. Nevertheless, when compounds are found to be effective in phenotypic assays, it’s often difficult to identify through which mechanism they are being effective.

While many techniques exists to perform target identification, they all have in common to be complex, time consuming, and even sometimes biased.

What does Zepto.Target do?

Zepto.Target, Kantify's solution for target deconvolution, is part of our AI drug discovery platform Zeptomics.

With Zepto.Target, we can predict:

What are the targets, or groups of targets, involved in a specific disease; What are the mechanisms of action through which a compound produces a particular effect.

Why is it unique?

Zepto.Target is a novel machine learning algorithm. Zepto.Target “deconvolutes” results from phenotypic drug screenings to identify likely targets and related Mechanisms of Action (MoA).

This revolutionary capability is helping us to quickly understand the targets and pathways of a disease, and accelerate in the optimization of lead compounds.

As it is pre-trained on millions of examples and data points, Zepto.Target can generate results very quickly.

Zepto.Target is a truly predictive, data-driven AI model. Because of its predictive nature, it is a less biased tool. This means that it does not rely on existing knowledge about a disease or a disease family to generate its predictions.

This makes it a unique solution in the field of target identification.

What are the benefits

The benefits of using Zepto.Target are the following:

Keep the benefits of phenotypic screenings without the uncertainty of target identification

Decrease the limits bias of existing available data

Considerably decrease time and cost of target identification efforts (from years to days);

Increase chances of success through the identification of the right target;

Speed up regulatory approval through the identification and the demonstration of the mechanism of action.

Discover Zepto.Target

Contact us by filling out the form below to learn more about using our target prediction solution Zepto.Target

Read about our AI for drug discovery platform Zeptomics on our platform page