Student or educator? Sign up with your school email and get 6 months of Pro free.
CoChat

The DOI matches its Crossref record; 156 of the 156 references it cites were found in a registry or an index.

DeepGAMI: deep biologically guided auxiliary learning for multimodal integration and imputation to improve genotype-phenotype prediction

Open access

Published version of a preprint

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.

Overview

The claims and statistics of this paper haven’t been checked yet.

Free: it reads the open-access text and checks the first 5 claims that cite a source we can check. Create a free account to check every claim.

A few free checks a day · papers already verified open instantly · results are public

Integrity signals

Registry

Crossref · 10.1186/s13073-023-01248-6

Cites

156 of 156 references verified0 citation errors · 0 not checkable

Statistics

Figures

Fig. 1 Architecture of DeepGAMI. A DeepGAMI first uses available multimodal features for training the predictive model, e.g., SNP genotypes (orange) and gene expression (blue) of individuals from the major applications in this study. In particular, it learns the latent space of each modality (e.g., consisting of latent features at the first transparent hidden layer). This learning step is also regularized by prior knowledge enabling biological interpretability after prediction, i.e., the input and latent features are connected by biological networks (biological DropConnect). For instance, the input transcription factor genes can be connected to target genes as their latent features (e.g., CGEX) by a gene regulatory network (GRN). The input SNPs can be connected to associated genes as their latent features (e.g., CSNP) by Expression quantitative trait loci (eQTLs). Notably, an auxiliary learning layer is used to infer the latent space of one modality to another, i.e., cross-modality imputation. For instance, DeepGAMI learns a transfer function f(.) to estimate CGEX from CSNP. Finally, the latent features are concatenated and fed to the feed-forward neural networks for phenotype predictions, e.g., classifying disease vs. control individuals. B Using the learned predictive model from multimodal input along, DeepGAMI can predict phenotypes from a single modality, e.g., SNP genotypes of new individuals. Specifically, it first imputes other modal latent spaces using the optimal transfer function f*(.) and then feeds both imputed and input latent features into downstream neural network predictions, e.g., C*GEX from CSNP
Fig. 1 Architecture of DeepGAMI. A DeepGAMI first uses available multimodal features for training the predictive model, e.g., SNP genotypes (orange) and gene expression (blue) of individuals from the major applications in this study. In particular, it learns the latent space of each modality (e.g., consisting of latent features at the first transparent hidden layer). This learning step is also regularized by prior knowledge enabling biological interpretability after prediction, i.e., the input and latent features are connected by biological networks (biological DropConnect). For instance, the input transcription factor genes can be connected to target genes as their latent features (e.g., CGEX) by a gene regulatory network (GRN). The input SNPs can be connected to associated genes as their latent features (e.g., CSNP) by Expression quantitative trait loci (eQTLs). Notably, an auxiliary learning layer is used to infer the latent space of one modality to another, i.e., cross-modality imputation. For instance, DeepGAMI learns a transfer function f(.) to estimate CGEX from CSNP. Finally, the latent features are concatenated and fed to the feed-forward neural networks for phenotype predictions, e.g., classifying disease vs. control individuals. B Using the learned predictive model from multimodal input along, DeepGAMI can predict phenotypes from a single modality, e.g., SNP genotypes of new individuals. Specifically, it first imputes other modal latent spaces using the optimal transfer function f*(.) and then feeds both imputed and input latent features into downstream neural network predictions, e.g., C*GEX from CSNPCC BY · PubMed Central
Fig. 2 Schizophrenia classification and functional genomic prioritization using genotype and bulk-tissue gene expression data. The population data was from the PsychENCODE project (Methods and materials). A Balanced accuracies from 5-fold cross-validation and B receiver operating characteristic (ROC) curves of DeepGAMI dual-modality model (dark blue), DeepGAMI single modality model (orange), Lasso (brown), LR (light blue), Random Forest (yellow), SVM (purple), Multilayer perceptron (MLP, pink), Varmole (red), and MOGONET (green) for classifying schizophrenia vs. control individuals on the held-out test samples. C ROC curves of various methods on cross-cohort SCZ prediction. D Select examples of prioritized transcription factors, SNPs, target genes (latent features, and functional links (GRNs, eQTLs) for schizophrenia. Purple: known schizophrenia genes. E Function and pathway enrichments of prioritized schizophrenia SNPs
Fig. 2 Schizophrenia classification and functional genomic prioritization using genotype and bulk-tissue gene expression data. The population data was from the PsychENCODE project (Methods and materials). A Balanced accuracies from 5-fold cross-validation and B receiver operating characteristic (ROC) curves of DeepGAMI dual-modality model (dark blue), DeepGAMI single modality model (orange), Lasso (brown), LR (light blue), Random Forest (yellow), SVM (purple), Multilayer perceptron (MLP, pink), Varmole (red), and MOGONET (green) for classifying schizophrenia vs. control individuals on the held-out test samples. C ROC curves of various methods on cross-cohort SCZ prediction. D Select examples of prioritized transcription factors, SNPs, target genes (latent features, and functional links (GRNs, eQTLs) for schizophrenia. Purple: known schizophrenia genes. E Function and pathway enrichments of prioritized schizophrenia SNPsCC BY · PubMed Central
Fig. 3 Multi-class clinical phenotype prediction and regulatory network prioritization in Alzheimer’s disease. A Fivefold cross-validation performance of DeepGAMI (Dual modality: dark blue, Single modality: orange) on three different phenotypes: neuritic plaque measure (CERAD score, multi-class), cognitive impairment (COGDX score, multi-class) and neurofibrillary tangle pathology (BRAAK stage, binary) in comparison with Lasso (brown), LR (light blue), Random Forest (yellow), SVM (purple), MLP (pink), and MOGONET (green). B ROC curves of held-out test samples for cognitive COGDX phenotype (blue) and late BRAAK stage (green). C Classification accuracies of the independent dataset for COGDX phenotype and BRAAK stage. D Enrichment analysis of prioritized genes for no cognitive impairment, mild cognitive impairment, and cognitive impairment (AD/Dementia) classes of COGDX phenotype. E Select an example of a prioritized regulatory network for the cognitive impairment phenotype. The edge thickness between any two nodes corresponds to the prioritized link importance score of the associated nodes. The edge color represents the three classes
Fig. 3 Multi-class clinical phenotype prediction and regulatory network prioritization in Alzheimer’s disease. A Fivefold cross-validation performance of DeepGAMI (Dual modality: dark blue, Single modality: orange) on three different phenotypes: neuritic plaque measure (CERAD score, multi-class), cognitive impairment (COGDX score, multi-class) and neurofibrillary tangle pathology (BRAAK stage, binary) in comparison with Lasso (brown), LR (light blue), Random Forest (yellow), SVM (purple), MLP (pink), and MOGONET (green). B ROC curves of held-out test samples for cognitive COGDX phenotype (blue) and late BRAAK stage (green). C Classification accuracies of the independent dataset for COGDX phenotype and BRAAK stage. D Enrichment analysis of prioritized genes for no cognitive impairment, mild cognitive impairment, and cognitive impairment (AD/Dementia) classes of COGDX phenotype. E Select an example of a prioritized regulatory network for the cognitive impairment phenotype. The edge thickness between any two nodes corresponds to the prioritized link importance score of the associated nodes. The edge color represents the three classesCC BY · PubMed Central
Fig. 4 Classifying cellular phenotype in single neuronal cells of mouse visual cortex. A DeepGAMI model for cell layer classification. B Balanced accuracies for classifying cell layers in the mouse visual cortex by DeepGAMI dual-modality (orange), DeepgGAMI single-modality (green) versus DeepManReg [92] (dark pink), neural network classification without any regularization (light pink), LMA [118] (violet), CCA [119] (blue), and MATCHEr [120] (brown). C Gene enrichment analysis showing the enriched terms for layer-specific prioritized genes
Fig. 4 Classifying cellular phenotype in single neuronal cells of mouse visual cortex. A DeepGAMI model for cell layer classification. B Balanced accuracies for classifying cell layers in the mouse visual cortex by DeepGAMI dual-modality (orange), DeepgGAMI single-modality (green) versus DeepManReg [92] (dark pink), neural network classification without any regularization (light pink), LMA [118] (violet), CCA [119] (blue), and MATCHEr [120] (brown). C Gene enrichment analysis showing the enriched terms for layer-specific prioritized genesCC BY · PubMed Central
Fig. 5 Classification of schizophrenia individuals and prioritization of genes, SNPs, and regulatory network using genotype and gene expressions of four cell types (microglia and astrocytes, oligodendrocytes, inhibitory neurons, and excitatory neurons). A Balanced accuracies from 5-fold cross-validation of DeepGAMI dual-modality model (dark blue), DeepGAMI single modality model (orange) in comparison with Lasso (brown), LR (light blue), Random Forest (yellow), SVM (purple), MLP (pink), Varmole (red), and MOGONET (green). B Pathway enrichment of prioritized schizophrenia SNPs for four cell types. The blue shade gives the − log(p-values) of the term. C Prioritized cell-type gene regulatory networks with pink circles representing schizophrenia genes. The size of the target gene is defined by the number of prioritized links between the SNPs and the associated gene
Fig. 5 Classification of schizophrenia individuals and prioritization of genes, SNPs, and regulatory network using genotype and gene expressions of four cell types (microglia and astrocytes, oligodendrocytes, inhibitory neurons, and excitatory neurons). A Balanced accuracies from 5-fold cross-validation of DeepGAMI dual-modality model (dark blue), DeepGAMI single modality model (orange) in comparison with Lasso (brown), LR (light blue), Random Forest (yellow), SVM (purple), MLP (pink), Varmole (red), and MOGONET (green). B Pathway enrichment of prioritized schizophrenia SNPs for four cell types. The blue shade gives the − log(p-values) of the term. C Prioritized cell-type gene regulatory networks with pink circles representing schizophrenia genes. The size of the target gene is defined by the number of prioritized links between the SNPs and the associated geneCC BY · PubMed Central

Verification

Claims the paper makes

The claims and statistics of this paper haven’t been checked yet.

Free: it reads the open-access text and checks the first 5 claims that cite a source we can check. Create a free account to check every claim.

A few free checks a day · papers already verified open instantly · results are public

Claims other papers make about this one

Each is checked against this paper’s own text: does it say what the citing paper says it does?

Self-check

Percentages, sums and reported differences re-computed from the paper’s own numbers.

The self-check needs the paper’s full text.

Citations

Cites

Show all 150 references

Research works printed but not linked: 5

  1. Sutton RS et al. Horde: a scalable real-time architecture for learning knowledge from unsupervised sensorimotor interaction. In: The 10th international conference on autonomous agents and multiagent systems-volume 2. 2011. p. 761–768.
  2. Lin X, Baweja H, Kantor G, Held D. Adaptive auxiliary task weighting for reinforcement learning. In: Proceedings of the 33rd Annual Conference on Neural Information Processing Systems (NIPS2019), Vancouver, BC, Canada; 2019.
  3. Paszke A, Gross S, Massa F, Lerer A, Bradbury J, Chanan G, Killeen T, Lin Z, Gimelshein N, Antiga L, Desmaison A. Pytorch: An imperative style, high-performance deep learning library. Adv Neural Inf Process Sys. 2019;32.
  4. Sundararajan M, Taly A, Yan Q. Axiomatic attribution for deep networks. In: Proceedings of the 34th International Conference on Machine Learning, Sydney, Australia. 2017;70:3319–28.
  5. Wang C, Mahadevan S. Alignment without correspondence. In: In Proceedings of the 21st International Joint Conferences on Artificial Intelligence. 2009.

Cited by

See all 41

Metadata

Authors
Chandrashekar, P. B., Alatkar, S., Wang, J., Hoffman, G. E., He, C., Jin, T., Khullar, S., Bendl, J., Fullard, J. F., Roussos, P., & Wang, D.
Affiliations
  • University of Wisconsin–Madison (US)
  • University of Pittsburgh (US)
  • Icahn School of Medicine at Mount Sinai (US)
  • Nathan Kline Institute for Psychiatric Research (US)
Corresponding author
Wang, Daifeng
Venue
Genome Medicine
Published
2023-10-31
Volume
15
Issue
1
Pages
88-88
ISSN
1756-994X (ISSN-L)
OpenAlex
W4388035454
Cited by
37
References
150
Citations by year
2026: 8 · 2025: 14 · 2024: 13 · 2023: 1
FWCI
5.98
Type
journal article
Language
en
Open access
Licence
Locations
  • Genome Medicine (publishedVersion, cc-by)
  • bioRxiv (Cold Spring Harbor Laboratory) (acceptedVersion)
  • PubMed Central (publishedVersion, cc-by)
  • PubMed (publishedVersion)
  • DOAJ (DOAJ: Directory of Open Access Journals) (submittedVersion, cc-by-sa)
Funders
  • National Institute on Aging (R01AG067025)
  • National Institute of Mental Health (RF1MH128695)
  • National Science Foundation (2144475)
Topics
  • Life Sciences › Biochemistry, Genetics and Molecular Biology › Genetics › Genetic Associations and Epidemiology
  • Life Sciences › Biochemistry, Genetics and Molecular Biology › Molecular Biology › Bioinformatics and Genomic Networks
  • Physical Sciences › Computer Science › Artificial Intelligence › Machine Learning in Healthcare
Keywords
genotype-phenotype prediction, multimodal data integration, auxiliary learning, interpretable neural networks, expression quantitative trait loci, gene expression regulation, integrated gradients, brain disorders, schizophrenia, Alzheimer's disease, cognitive impairment, electrophysiology, disease-associated variants, regulatory networks
MeSH terms
Machine Learning, Alzheimer Disease, Animals, Genotype, Phenotype, Neural Networks, Computer, Mice

About these checks

Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.

Statistics and arithmetic are recomputed, not judged.

Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.

What counts as a claim ›How we measure confidence ›How we check this ›

Are you an author, or do you see a problem with this page? Request a review

How to cite

Cite CoChat for the checks on this page, and the paper for its own findings.

This page
CoChat. (2026). Citation check: DeepGAMI: deep biologically guided auxiliary learning for multimodal integration and imputation to improve genotype-phenotype prediction. https://app.cochat.ai/papers/10.1186/s13073-023-01248-6
The paper
Chandrashekar, P. B., Alatkar, S., Wang, J., Hoffman, G. E., He, C., Jin, T., Khullar, S., Bendl, J., Fullard, J. F., Roussos, P., & Wang, D. (2023). DeepGAMI: deep biologically guided auxiliary learning for multimodal integration and imputation to improve genotype-phenotype prediction. Genome Medicine. https://doi.org/10.1186/s13073-023-01248-6

Data sources

Bibliographic data: Semantic Scholar, OpenAlex, Crossref.

Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).

Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.