DeepGAMI: deep biologically guided auxiliary learning for multimodal integration and imputation to improve genotype–phenotype prediction
Abstract Background Genotypes are strongly associated with disease phenotypes, particularly in brain disorders. However, the molecular and cellular mechanisms behind this association remain elusive. With emerging multimodal data for these mechanisms, machine learning methods can be applied for pheno...
Main Authors: | , , , , , , , , , , |
---|---|
Format: | Article |
Language: | English |
Published: |
BMC
2023-10-01
|
Series: | Genome Medicine |
Subjects: | |
Online Access: | https://doi.org/10.1186/s13073-023-01248-6 |
_version_ | 1797636777779920896 |
---|---|
author | Pramod Bharadwaj Chandrashekar Sayali Alatkar Jiebiao Wang Gabriel E. Hoffman Chenfeng He Ting Jin Saniya Khullar Jaroslav Bendl John F. Fullard Panos Roussos Daifeng Wang |
author_facet | Pramod Bharadwaj Chandrashekar Sayali Alatkar Jiebiao Wang Gabriel E. Hoffman Chenfeng He Ting Jin Saniya Khullar Jaroslav Bendl John F. Fullard Panos Roussos Daifeng Wang |
author_sort | Pramod Bharadwaj Chandrashekar |
collection | DOAJ |
description | Abstract Background Genotypes are strongly associated with disease phenotypes, particularly in brain disorders. However, the molecular and cellular mechanisms behind this association remain elusive. With emerging multimodal data for these mechanisms, machine learning methods can be applied for phenotype prediction at different scales, but due to the black-box nature of machine learning, integrating these modalities and interpreting biological mechanisms can be challenging. Additionally, the partial availability of these multimodal data presents a challenge in developing these predictive models. Method To address these challenges, we developed DeepGAMI, an interpretable neural network model to improve genotype–phenotype prediction from multimodal data. DeepGAMI leverages functional genomic information, such as eQTLs and gene regulation, to guide neural network connections. Additionally, it includes an auxiliary learning layer for cross-modal imputation allowing the imputation of latent features of missing modalities and thus predicting phenotypes from a single modality. Finally, DeepGAMI uses integrated gradient to prioritize multimodal features for various phenotypes. Results We applied DeepGAMI to several multimodal datasets including genotype and bulk and cell-type gene expression data in brain diseases, and gene expression and electrophysiology data of mouse neuronal cells. Using cross-validation and independent validation, DeepGAMI outperformed existing methods for classifying disease types, and cellular and clinical phenotypes, even using single modalities (e.g., AUC score of 0.79 for Schizophrenia and 0.73 for cognitive impairment in Alzheimer’s disease). Conclusion We demonstrated that DeepGAMI improves phenotype prediction and prioritizes phenotypic features and networks in multiple multimodal datasets in complex brains and brain diseases. Also, it prioritized disease-associated variants, genes, and regulatory networks linked to different phenotypes, providing novel insights into the interpretation of gene regulatory mechanisms. DeepGAMI is open-source and available for general use. |
first_indexed | 2024-03-11T12:39:02Z |
format | Article |
id | doaj.art-c8c7e6c0ecce42439e257b1b9cc4a68e |
institution | Directory Open Access Journal |
issn | 1756-994X |
language | English |
last_indexed | 2024-03-11T12:39:02Z |
publishDate | 2023-10-01 |
publisher | BMC |
record_format | Article |
series | Genome Medicine |
spelling | doaj.art-c8c7e6c0ecce42439e257b1b9cc4a68e2023-11-05T12:25:43ZengBMCGenome Medicine1756-994X2023-10-0115111910.1186/s13073-023-01248-6DeepGAMI: deep biologically guided auxiliary learning for multimodal integration and imputation to improve genotype–phenotype predictionPramod Bharadwaj Chandrashekar0Sayali Alatkar1Jiebiao Wang2Gabriel E. Hoffman3Chenfeng He4Ting Jin5Saniya Khullar6Jaroslav Bendl7John F. Fullard8Panos Roussos9Daifeng Wang10Waisman Center, University of Wisconsin-MadisonWaisman Center, University of Wisconsin-MadisonDepartment of Biostatistics, University of PittsburghCenter for Disease Neurogenomics, Department of Psychiatry and Department of Genetics and Genomic Sciences, Icahn School of Medicine at Mount SinaiWaisman Center, University of Wisconsin-MadisonWaisman Center, University of Wisconsin-MadisonWaisman Center, University of Wisconsin-MadisonCenter for Disease Neurogenomics, Department of Psychiatry and Department of Genetics and Genomic Sciences, Icahn School of Medicine at Mount SinaiCenter for Disease Neurogenomics, Department of Psychiatry and Department of Genetics and Genomic Sciences, Icahn School of Medicine at Mount SinaiCenter for Disease Neurogenomics, Department of Psychiatry and Department of Genetics and Genomic Sciences, Icahn School of Medicine at Mount SinaiWaisman Center, University of Wisconsin-MadisonAbstract Background Genotypes are strongly associated with disease phenotypes, particularly in brain disorders. However, the molecular and cellular mechanisms behind this association remain elusive. With emerging multimodal data for these mechanisms, machine learning methods can be applied for phenotype prediction at different scales, but due to the black-box nature of machine learning, integrating these modalities and interpreting biological mechanisms can be challenging. Additionally, the partial availability of these multimodal data presents a challenge in developing these predictive models. Method To address these challenges, we developed DeepGAMI, an interpretable neural network model to improve genotype–phenotype prediction from multimodal data. DeepGAMI leverages functional genomic information, such as eQTLs and gene regulation, to guide neural network connections. Additionally, it includes an auxiliary learning layer for cross-modal imputation allowing the imputation of latent features of missing modalities and thus predicting phenotypes from a single modality. Finally, DeepGAMI uses integrated gradient to prioritize multimodal features for various phenotypes. Results We applied DeepGAMI to several multimodal datasets including genotype and bulk and cell-type gene expression data in brain diseases, and gene expression and electrophysiology data of mouse neuronal cells. Using cross-validation and independent validation, DeepGAMI outperformed existing methods for classifying disease types, and cellular and clinical phenotypes, even using single modalities (e.g., AUC score of 0.79 for Schizophrenia and 0.73 for cognitive impairment in Alzheimer’s disease). Conclusion We demonstrated that DeepGAMI improves phenotype prediction and prioritizes phenotypic features and networks in multiple multimodal datasets in complex brains and brain diseases. Also, it prioritized disease-associated variants, genes, and regulatory networks linked to different phenotypes, providing novel insights into the interpretation of gene regulatory mechanisms. DeepGAMI is open-source and available for general use.https://doi.org/10.1186/s13073-023-01248-6Deep learningCross-modality imputationAuxiliary learningGenotype–phenotype predictionCell-type gene regulatory networksSchizophrenia |
spellingShingle | Pramod Bharadwaj Chandrashekar Sayali Alatkar Jiebiao Wang Gabriel E. Hoffman Chenfeng He Ting Jin Saniya Khullar Jaroslav Bendl John F. Fullard Panos Roussos Daifeng Wang DeepGAMI: deep biologically guided auxiliary learning for multimodal integration and imputation to improve genotype–phenotype prediction Genome Medicine Deep learning Cross-modality imputation Auxiliary learning Genotype–phenotype prediction Cell-type gene regulatory networks Schizophrenia |
title | DeepGAMI: deep biologically guided auxiliary learning for multimodal integration and imputation to improve genotype–phenotype prediction |
title_full | DeepGAMI: deep biologically guided auxiliary learning for multimodal integration and imputation to improve genotype–phenotype prediction |
title_fullStr | DeepGAMI: deep biologically guided auxiliary learning for multimodal integration and imputation to improve genotype–phenotype prediction |
title_full_unstemmed | DeepGAMI: deep biologically guided auxiliary learning for multimodal integration and imputation to improve genotype–phenotype prediction |
title_short | DeepGAMI: deep biologically guided auxiliary learning for multimodal integration and imputation to improve genotype–phenotype prediction |
title_sort | deepgami deep biologically guided auxiliary learning for multimodal integration and imputation to improve genotype phenotype prediction |
topic | Deep learning Cross-modality imputation Auxiliary learning Genotype–phenotype prediction Cell-type gene regulatory networks Schizophrenia |
url | https://doi.org/10.1186/s13073-023-01248-6 |
work_keys_str_mv | AT pramodbharadwajchandrashekar deepgamideepbiologicallyguidedauxiliarylearningformultimodalintegrationandimputationtoimprovegenotypephenotypeprediction AT sayalialatkar deepgamideepbiologicallyguidedauxiliarylearningformultimodalintegrationandimputationtoimprovegenotypephenotypeprediction AT jiebiaowang deepgamideepbiologicallyguidedauxiliarylearningformultimodalintegrationandimputationtoimprovegenotypephenotypeprediction AT gabrielehoffman deepgamideepbiologicallyguidedauxiliarylearningformultimodalintegrationandimputationtoimprovegenotypephenotypeprediction AT chenfenghe deepgamideepbiologicallyguidedauxiliarylearningformultimodalintegrationandimputationtoimprovegenotypephenotypeprediction AT tingjin deepgamideepbiologicallyguidedauxiliarylearningformultimodalintegrationandimputationtoimprovegenotypephenotypeprediction AT saniyakhullar deepgamideepbiologicallyguidedauxiliarylearningformultimodalintegrationandimputationtoimprovegenotypephenotypeprediction AT jaroslavbendl deepgamideepbiologicallyguidedauxiliarylearningformultimodalintegrationandimputationtoimprovegenotypephenotypeprediction AT johnffullard deepgamideepbiologicallyguidedauxiliarylearningformultimodalintegrationandimputationtoimprovegenotypephenotypeprediction AT panosroussos deepgamideepbiologicallyguidedauxiliarylearningformultimodalintegrationandimputationtoimprovegenotypephenotypeprediction AT daifengwang deepgamideepbiologicallyguidedauxiliarylearningformultimodalintegrationandimputationtoimprovegenotypephenotypeprediction |