2021
Ancestry may confound genetic machine learning: Candidate-gene prediction of opioid use disorder as an example
Hatoum AS, Wendt FR, Galimberti M, Polimanti R, Neale B, Kranzler HR, Gelernter J, Edenberg HJ, Agrawal A. Ancestry may confound genetic machine learning: Candidate-gene prediction of opioid use disorder as an example. Drug And Alcohol Dependence 2021, 229: 109115. PMID: 34710714, PMCID: PMC9358969, DOI: 10.1016/j.drugalcdep.2021.109115.Peer-Reviewed Original ResearchMeSH KeywordsBlack PeopleHumansMachine LearningMultifactorial InheritanceOpioid-Related DisordersPolymorphism, Single NucleotideConceptsGenome-wide significant variantsCandidate gene predictionGenetic predictionRandom SNPsPolygenic traitRandom phenotypeCandidate SNPsSimulated phenotypesPsychiatric geneticsGenetic machineSignificant variantsBinary phenotypesCandidate variantsSNPsAncestryPhenotypeAllele frequenciesVariantsMachine learning modelsGenetic testsLearning model
2018
Multivariate Pattern Analysis of Genotype–Phenotype Relationships in Schizophrenia
Zheutlin AB, Chekroud AM, Polimanti R, Gelernter J, Sabb FW, Bilder RM, Freimer N, London ED, Hultman CM, Cannon TD. Multivariate Pattern Analysis of Genotype–Phenotype Relationships in Schizophrenia. Schizophrenia Bulletin 2018, 44: 1045-1052. PMID: 29534239, PMCID: PMC6101611, DOI: 10.1093/schbul/sby005.Peer-Reviewed Original ResearchConceptsMultivariate pattern analysisIndependent samplesVisual memoryCognitive endophenotypesPredictive strengthSchizophreniaMemoryIndividual variationPattern analysisSingle predictorCertain domainsDiscovery samplePsychiatric patientsPolygenic risk scoresPredictive powerScoresEndophenotypesPotential relationshipRelationshipRandom forestGenetic risk variantsLimited setPredictorsComprehensive setSamples