Search research articles

ABOUT JoVE

Overview Leadership Blog JoVE Help Center

AUTHORS

Publishing Process Editorial Board Scope & Policies Peer Review FAQ Submit

LIBRARIANS

Testimonials Subscriptions Access Resources Library Advisory Board FAQ

RESEARCH

JoVE Journal Methods Collections JoVE Encyclopedia of Experiments Archive

EDUCATION

JoVE Core JoVE Business JoVE Science Education JoVE Lab Manual Faculty Resource Center Faculty Site

Terms & Conditions of Use

Related Concept Videos

Pharmacogenomics: Identification of New Drug Targets

Pharmacogenomics: Identification of New Drug Targets

Advances in genomics have profoundly influenced drug discovery by increasing both the speed and accuracy of pharmaceutical development. Pharmacogenomics, which examines how genetic variation influences drug response, facilitates the identification of novel therapeutic targets and enables patient stratification for personalized treatment. These strategies contribute to improved drug efficacy, minimized adverse effects, and more efficient clinical trial design.Mapping genetic differences...

You might also read

Related Articles

Articles linked to this work by shared authors, journal, and citation graph.

Sort by

Same author

CDKL5 deficiency results in atypical subregion-specific expression of perineuronal nets during mouse visual critical period.

Frontiers in neuroscience·2026

Same author

Mechanisms of Binding and Immune Escape Resistance for Broadly Neutralizing Antibodies Targeting Distinct Conserved SARS-CoV-2 Spike Epitopes: A Hierarchical Approach Integrating Mutational Profiling and Energy Landscape Analysis.

International journal of molecular sciences·2026

Same author

Predicting and Decoding Allosteric Binding Sites Using Protein Language Models and Structure-Based Machine Learning: An Energy Landscape-Guided Explainable AI Framework.

Journal of chemical theory and computation·2026

Same author

Frustration Landscapes of Broadly Neutralizing SARS-CoV-2 Spike Antibodies Targeting Conserved Epitopes Reveal Energetic Logic of Escape-Proof and Escape-Prone Mechanisms.

bioRxiv : the preprint server for biology·2026

Same author

Decoding Allosteric Grammar with Explainable AI Integrating Protein Language Models and Energy Landscape Analysis: Neutral Frustration at Allosteric Binding Sites Encodes Regulatory Versatility in Protein Kinases.

bioRxiv : the preprint server for biology·2026

Same author

Decoding the Allosteric Paradox: A Dual Framework Integrating AI Cofolding Models with Landscape-Guided Interpretable AI Framework of Ligand-Protein Binding.

bioRxiv : the preprint server for biology·2026

Search research articles

Related Experiment Video

Updated: Feb 28, 2026

Characterization at the Molecular Level using Robust Biochemical Approaches of a New Kinase Protein

Characterization at the Molecular Level using Robust Biochemical Approaches of a New Kinase Protein

Published on: June 30, 2019

From Latent Manifolds to Targeted Molecular Probes: An Interpretable, Kinome-Scale Generative Machine Learning

Gennady Verkhivker^1,2, Ryan Kassab¹, Keerthi Krishnan¹

¹Graduate Program in Computational and Data Sciences, Keck Center for Science and Engineering, Schmid College of Science and Technology, Chapman University, Orange, CA 92866, USA.

|February 27, 2026

Summary

This study introduces a new AI framework for designing kinase ligands, revealing how molecular structure influences AI models. It highlights limitations in current AI representations for complex molecules like kinase inhibitors.

Keywords:

autonomous molecular design deep learning models explainable machine learning kinase association likelihood classifiers kinase ligands latent space landscapes local neighborhood sampling chemical modeling protein kinases

More Related Videos

Kinase Inhibitor Screening In Self-assembled Human Protein Microarrays

Kinase Inhibitor Screening In Self-assembled Human Protein Microarrays

Published on: October 23, 2019

Identification of Kinase-substrate Pairs Using High Throughput Screening

Identification of Kinase-substrate Pairs Using High Throughput Screening

Published on: August 29, 2015

Related Experiment Videos

Last Updated: Feb 28, 2026

Characterization at the Molecular Level using Robust Biochemical Approaches of a New Kinase Protein

Characterization at the Molecular Level using Robust Biochemical Approaches of a New Kinase Protein

Published on: June 30, 2019

Kinase Inhibitor Screening In Self-assembled Human Protein Microarrays

Kinase Inhibitor Screening In Self-assembled Human Protein Microarrays

Published on: October 23, 2019

Identification of Kinase-substrate Pairs Using High Throughput Screening

Identification of Kinase-substrate Pairs Using High Throughput Screening

Published on: August 29, 2015

Area of Science:

Computational chemistry
Artificial intelligence in drug discovery
Molecular modeling

Background:

Scaffold-aware AI models are crucial for exploring chemical space but their representational principles are unclear.
Representing complex kinase small molecules is challenging due to conserved ATP active sites and scaffold complexity in generative AI.
Understanding AI representational limits is key for designing targeted kinase ligands.

Purpose of the Study:

To develop and evaluate a diagnostic, modular, chemistry-first generative AI framework for SRC kinase ligand design.
To investigate scaffold topology, latent-space geometry, and generative trajectories using a kinase ligand dataset.
To address the limitations of current molecular representations in AI for kinase inhibitor design.

Main Methods:

Integrated ChemVAE latent space modeling with a chemically interpretable similarity metric (Kinase Likelihood Score).
Employed Bayesian optimization and cluster-guided local neighborhood sampling for scaffold transformation.
Analyzed scaffold topology and latent-space geometry across 37 protein kinase families.

Main Results:

Chemically distinct scaffolds can have overlapping latent representations, indicating encoding degeneracy.
Specific topological motifs act as anchors, constraining generative diversification.
SRC-like scaffolds function as a structural hub, enabling rational scaffold transformation and conversion of LCK scaffolds to SRC-like chemotypes.
SMILES-based representations fail to capture multi-ring aromatic systems, a key feature of kinase chemotypes, creating a 'representation gap'.

Conclusions:

The developed framework provides a conceptual basis for interpreting generative AI behavior in ligand design.
Molecular representation is critical; AI must access topologically constrained regions for effective kinase inhibitor design.
Incorporating structural priors into AI architectures is essential for overcoming current representational limitations.