Quantitative structure-activity relationship analysis using deep learning based on a novel molecular image input technique.

Yoshihiro Uesawa
Author Information
  1. Yoshihiro Uesawa: Department of Medical Molecular Informatics, Meiji Pharmaceutical University, 2-522-1 Noshio, Kiyose, Tokyo 204-8588, Japan. Electronic address: uesawa@my-pharm.ac.jp.

Abstract

Quantitative structure-activity relationship (QSAR) analysis uses structural, quantum chemical, and physicochemical features calculated from molecular geometry as explanatory variables predicting physiological activity. Recently, deep learning based on advanced artificial neural networks has demonstrated excellent performance in the discipline of QSAR research. While it has properties of feature representation learning that directly calculate feature values from molecular structure, the use of this potential function is limited in QSAR modeling. The present study applied this function of feature representation learning to QSAR analysis by incorporating 360° images of molecular conformations into deep learning. Accordingly, I successfully constructed a highly versatile identification model for chemical compounds that induce mitochondrial membrane potential disruption with the external validation area under the receiver operating characteristic curve of ≥0.9.

Keywords

MeSH Term

Deep Learning
Models, Chemical
Molecular Conformation
Molecular Structure
Quantitative Structure-Activity Relationship
ROC Curve

Word Cloud

Created with Highcharts 10.0.0learningQSARmolecularQuantitativerelationshipanalysisdeepfeaturepotentialstructure-activitychemicalbasedrepresentationstructurefunctionmembranedisruptionusesstructuralquantumphysicochemicalfeaturescalculatedgeometryexplanatoryvariablespredictingphysiologicalactivityRecentlyadvancedartificialneuralnetworksdemonstratedexcellentperformancedisciplineresearchpropertiesdirectlycalculatevaluesuselimitedmodelingpresentstudyappliedincorporating360°imagesconformationsAccordinglysuccessfullyconstructedhighlyversatileidentificationmodelcompoundsinducemitochondrialexternalvalidationareareceiveroperatingcharacteristiccurve≥09usingnovelimageinputtechniqueDeepsilicoMitochondrialMolecularimagerystructure–activityThree-dimensional

Similar Articles

Cited By