2021
PaccMannRL: De novo generation of hit-like anticancer molecules from transcriptomic data via reinforcement learning
Born J, Manica M, Oskooei A, Cadow J, Markert G, Martínez M. PaccMannRL: De novo generation of hit-like anticancer molecules from transcriptomic data via reinforcement learning. IScience 2021, 24: 102269. PMID: 33851095, PMCID: PMC8022157, DOI: 10.1016/j.isci.2021.102269.Peer-Reviewed Original ResearchDrug designSimilarity to compoundsReinforcement learning methodDeep generative modelsDeep learning approachMolecular designStructurally similar to compoundsDrug-likenessComputational chemistryBridging systems biologyMolecule generationReinforcement learningReward functionLearning methodsGenerative modelLearning approachCompoundsAnticancer moleculesChemical propertiesMoleculesIncorporating informationCandidate drugsSynthesizabilityAnticancer drugsPrediction model
2019
Toward Explainable Anticancer Compound Sensitivity Prediction via Multimodal Attention-Based Convolutional Encoders
Manica M, Oskooei A, Born J, Subramanian V, Sáez-Rodríguez J, Martínez M. Toward Explainable Anticancer Compound Sensitivity Prediction via Multimodal Attention-Based Convolutional Encoders. Molecular Pharmaceutics 2019, 16: 4797-4806. PMID: 31618586, DOI: 10.1021/acs.molpharmaceut.9b00520.Peer-Reviewed Original ResearchConceptsConvolutional encoderReceptor tyrosine kinasesProtein-protein interaction networkAttention-based encoderStructural similarity indexSelection of encodingDrug designDrug sensitivity predictionGene expression profilesIn silico predictionSensitivity predictionAttention weightsLeukemia cell linesSMILES sequencesInformative genesGene expression profiles of tumorsApoptotic processInteraction networkExpression profiles of tumorsBaseline modelIntracellular interactionsEncodingTyrosine kinaseDevelopment of personalized therapiesGenes