Cheminformatics

Achieve compounds with favorable potency, selectivity, and ADME profile by leveraging advanced computational modeling and state-of-the-art artificial intelligence. 

Our proven expertise has been demonstrated by successful client projects, strong publication records, and top performance in global AI/ML competitions. 

We deliver detailed analyses, regular project updates, prioritized compound lists, and actionable insights, all tailored to your project’s needs. 

Our cheminformatics team covers different areas of expertise, including:

Target structure validation 

• Ligandability assessment: Ensure reasonable starting point for structure-guided drug design 

• Homology modeling: Build 3D model of the target from a template  

• Cryptic pocket detection: Identify potential pockets not visible in PDB structures using molecular dynamics 

Hit identification and expansion 

• Docking & pharmacophoric modeling: Predict ligand pose and identify key interactions 

• Virtual screening: Identify hits from vast synthesizable commercial libraries by high throughput docking 

• Molecular dynamics simulation: Study ligand stability, impact of water interactions, and induced-fit behavior 

Lead generation and optimization 

• Potency and ADME prediction: Build tailored predictive models from SAR & ADME data 

• AI growing and scaffold hopping: Improve potency of initial hits by AI-driven chemical space exploration 

• AI for patent scope expansion: Explore all potential isosteric replacements of your lead series, reducing the risk of patent-busting 
  

Molecular Dynamics trajectory by cheminformatics team on cryo-EM structure released internally by Structural Biology and cryo-EM teams.

AI-driven cheminformatics recognized for excellence

Our in-house machine learning pipelines achieved top international rankings in the Antiviral Competition organized by ASAP Discovery, OpenADMET, and Polaris.
Among more than 60 participating teams, our models ranked #1 for IC₅₀ prediction on the SARS-CoV-2 Mpro target and #4 for early ADME prediction, demonstrating the robustness and innovation of our AI-driven drug discovery approaches.

Learn more about this achievement in our LinkedIn post.

Discover our team’s publications

• Evaluating the Diversity and Target Addressability of DELs using Scaffold Analysis and Machine Learning

• Deep learning vs classical methods in potency & ADME prediction: Insights from the Polaris Antiviral Challenge

• A Computational Community Blind Challenge on Pan-Coronavirus Drug Discovery Data

• Combining DEL and AI for hit finding: hype or reality?

Explore how our computational chemistry expertise can advance your projects — contact us to learn more.