AI-ACCELERATED DRUG DISCOVERY
Improved performance in predicting ligand binding poses and accounting for ions and cofactors
We are honored to announce the release of ArtiDock 2.5, a significantly upgraded version of the molecular docking model developed by Receptor.AI. This version provides enhanced capabilities for predicting the binding poses of small molecule ligands in different pocket representations accounting for the non-standard residues, ions, and co-factors, surpassing classic docking techniques. Now seamlessly integrated into our Drug Design Ecosystem, ArtiDock 2.5 is ready to multiply R&D throughput.
Key upgrades: Simulating Binding Site Variability
The main advancement of ArtiDock 2.5 is the employment of augmented pocket representations during the training.
In the real-world scenario, the researcher doesn’t know the true ligand pose in the protein pocket, otherwise, there would be no need to perform docking. Therefore, a significant variability of potential pocket representations could be selected before the docking.
Our solution to this issue is the introduction of random binding pocket extraction in the training process. This method simulates possible variations of the binding sites that the researcher might use as input. This approach allowed us to increase the percentage of correct predictions up to 2-fold in the pockets of a different size (Fig. 1).
Accounting for ions and cofactors
Another upgrade in ArtiDock 2.5 is the accounting for modified or non-standard residues, ions (Fig. 2), and cofactors (Fig. 3). We modified the training set to include all possible not-protein entities except water, which led to improved model performance in a variety of cases.
The researcher may now include any molecular structures in the binding pocket and ArtiDock will handle them.
These updates improve ArtiDock's performance in a variety of real-world scenarios, enhancing the impact of the technology within a drug discovery setting.
ArtiDock 2.5 is fully integrated into our ecosystem and ready for use. We look forward to receiving your feedback.
