Jove
Visualize
Contact Us
JoVE
x logofacebook logolinkedin logoyoutube logo
ABOUT JoVE
OverviewLeadershipBlogJoVE Help Center
AUTHORS
Publishing ProcessEditorial BoardScope & PoliciesPeer ReviewFAQSubmit
LIBRARIANS
TestimonialsSubscriptionsAccessResourcesLibrary Advisory BoardFAQ
RESEARCH
JoVE JournalMethods CollectionsJoVE Encyclopedia of ExperimentsArchive
EDUCATION
JoVE CoreJoVE BusinessJoVE Science EducationJoVE Lab ManualFaculty Resource CenterFaculty Site
Terms & Conditions of Use
Privacy Policy
Policies

Related Experiment Videos

Automatic generation of complementary descriptors with molecular graph networks.

Christian Merkwirth1, Thomas Lengauer

  • 1Computational Biology and Applied Algorithmics Group, Max-Planck-Institut für Informatik, Stuhlsatzenhausweg 85, 66123 Saarbrücken, Germany. cmerk@mpi-sb.mpg.de

Journal of Chemical Information and Modeling
|September 27, 2005
PubMed
Summary

This study introduces a novel method for generating molecular descriptors from chemical structures. The approach uses statistical learning to create adaptive descriptors, showing competitive performance in AIDS antiviral drug screening.

Related Concept Videos

You might also read

Related Articles

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

Sort by
Same author

Geno2pheno: recombination detection for HIV-1 and HEV subtypes.

NAR molecular medicine·2025
Same author

Biological databases in the age of generative artificial intelligence.

Bioinformatics advances·2025
Same author

Hepatitis C Virus Saint Petersburg Variant Detection With Machine Learning Methods.

Journal of medical virology·2025
Same author

Current methods for detecting and assessing HIV-1 antibody resistance.

Frontiers in immunology·2025
Same author

Incorporating temporal dynamics of mutations to enhance the prediction capability of antiretroviral therapy's outcome for HIV-1.

Bioinformatics (Oxford, England)·2024
Same author

Yves Moreau has received the 2023 Einstein Foundation Individual Award for Promoting Quality in Research.

Bioinformatics advances·2024

Area of Science:

  • Computational chemistry
  • Chemoinformatics
  • Machine learning

Background:

  • Developing effective molecular descriptors is crucial for drug discovery and chemical data analysis.
  • Existing methods may generate highly correlated descriptors, limiting their utility.
  • Automatic generation of robust descriptors is needed for large-scale screening.

Purpose of the Study:

  • To present a new method for automatically generating weakly correlated molecular descriptors.
  • To adapt molecular graph structures into a composite descriptor using statistical learning.
  • To evaluate the method's performance on a real-world drug screening dataset.

Main Methods:

  • The method translates molecular graphs into dynamical systems for descriptor generation.

Related Experiment Videos

  • It employs gradient descent and back-propagation for training the descriptor generation system.
  • Descriptors are designed to be sensitive to molecular topology.

    • Main Results:

    • The developed method successfully generates adaptive, whole-molecule composite descriptors.
    • Computational experiments demonstrated the method's effectiveness in classifying the Developmental Therapeutics Program AIDS antiviral screen dataset.
    • Performance was comparable to established substructure comparison methods.

    Conclusions:

    • The proposed method offers an efficient way to generate weakly correlated molecular descriptors.
    • This approach holds promise for improving virtual screening and drug discovery pipelines.
    • The technique provides a valuable tool for chemoinformatics and computational drug design.