Applying interpretable machine learning in computational biology-pitfalls, recommendations and opportunities for new developments.


Journal

Nature methods
ISSN: 1548-7105
Titre abrégé: Nat Methods
Pays: United States
ID NLM: 101215604

Informations de publication

Date de publication:
Aug 2024
Historique:
received: 27 10 2022
accepted: 24 06 2024
medline: 10 8 2024
pubmed: 10 8 2024
entrez: 9 8 2024
Statut: ppublish

Résumé

Recent advances in machine learning have enabled the development of next-generation predictive models for complex computational biology problems, thereby spurring the use of interpretable machine learning (IML) to unveil biological insights. However, guidelines for using IML in computational biology are generally underdeveloped. We provide an overview of IML methods and evaluation techniques and discuss common pitfalls encountered when applying IML methods to computational biology problems. We also highlight open questions, especially in the era of large language models, and call for collaboration between IML and computational biology researchers.

Identifiants

pubmed: 39122941
doi: 10.1038/s41592-024-02359-7
pii: 10.1038/s41592-024-02359-7
doi:

Types de publication

Journal Article Review

Langues

eng

Sous-ensembles de citation

IM

Pagination

1454-1461

Subventions

Organisme : U.S. Department of Health & Human Services | National Institutes of Health (NIH)
ID : R01HG012303
Organisme : U.S. Department of Health & Human Services | National Institutes of Health (NIH)
ID : R01HG007352
Organisme : U.S. Department of Health & Human Services | National Institutes of Health (NIH)
ID : UM1HG011593
Organisme : National Science Foundation (NSF)
ID : IIS1705121
Organisme : National Science Foundation (NSF)
ID : IIS1838017
Organisme : National Science Foundation (NSF)
ID : IIS2046613
Organisme : National Science Foundation (NSF)
ID : IIS2112471

Informations de copyright

© 2024. Springer Nature America, Inc.

Références

Miller, T. Explanation in artificial intelligence: insights from the social sciences. Artif. Intell. 267, 1–38 (2019).
doi: 10.1016/j.artint.2018.07.007
Doshi-Velez, F. & Kim, B. Towards a rigorous science of interpretable machine learning. Preprint at https://arxiv.org/abs/1702.08608 (2017).
Azodi, C. B., Tang, J. & Shiu, S. -H. Opening the black box: interpretable machine learning for geneticists. Trends Genet. 36, 442–455 (2020).
pubmed: 32396837 doi: 10.1016/j.tig.2020.03.005
Eraslan, G., Avsec, Ž., Gagneur, J. & Theis, F. J. Deep learning: new computational modelling techniques for genomics. Nat. Rev. Genet. 20, 389–403 (2019). This paper gives an extensive review of the application of deep learning models in genomics.
pubmed: 30971806 doi: 10.1038/s41576-019-0122-6
Talukder, A., Barham, C., Li, X. & Hu, H. Interpretation of deep learning in genomics and epigenomics. Brief. Bioinform. 22, bbaa177 (2021).
pubmed: 34020542 doi: 10.1093/bib/bbaa177
Novakovsky, G., Dexter, N., Libbrecht, M. W., Wasserman, W. W. & Mostafavi, S. Obtaining genetics insights from deep learning via explainable artificial intelligence. Nat. Rev. Genet. 24, 125–137 (2023). This paper provides a comprehensive review for the commonly applied IML methods in biology through the examples from regulatory genomics.
pubmed: 36192604 doi: 10.1038/s41576-022-00532-2
Klauschen, F. et al. Toward explainable artificial intelligence for precision pathology. Annu. Rev. Pathol. 19, 541–570 (2024).
pubmed: 37871132 doi: 10.1146/annurev-pathmechdis-051222-113147
Avsec, Ž. et al. Effective gene expression prediction from sequence by integrating long-range interactions. Nat. Methods 18, 1196–1203 (2021).
pubmed: 34608324 pmcid: 8490152 doi: 10.1038/s41592-021-01252-x
Avsec, Ž. et al. Base-resolution models of transcription-factor binding reveal soft motif syntax. Nat. Genet. 53, 354–366 (2021). This paper is a representative example of applying post hoc explanation methods and connecting the feature importance scores with biological interpretations.
pubmed: 33603233 pmcid: 8812996 doi: 10.1038/s41588-021-00782-6
Schwessinger, R. et al. DeepC: predicting 3D genome folding using megabase-scale transfer learning. Nat. Methods 17, 1118–1124 (2020).
pubmed: 33046896 pmcid: 7610627 doi: 10.1038/s41592-020-0960-3
Karimi, M., Wu, D., Wang, Z. & Shen, Y. DeepAffinity: interpretable deep learning of compound–protein affinity through unified recurrent and convolutional neural networks. Bioinformatics 35, 3329–3338 (2019).
pubmed: 30768156 pmcid: 6748780 doi: 10.1093/bioinformatics/btz111
Vig, J. et al. BERTology meets biology: interpreting attention in protein language models. In International Conference on Learning Representations (ICLR, 2021).
Taujale, R. et al. Mapping the glycosyltransferase fold landscape using interpretable deep learning. Nat. Commun. 12, 5656 (2021).
pubmed: 34580305 pmcid: 8476585 doi: 10.1038/s41467-021-25975-9
Ma, J. et al. Using deep learning to model the hierarchical structure and function of a cell. Nat. Methods 15, 290–298 (2018). This paper illustrates a biologically informed neural network that incorporates the hierarchical cell subsystems into the neural network architecture.
pubmed: 29505029 pmcid: 5882547 doi: 10.1038/nmeth.4627
Tasaki, S., Gaiteri, C., Mostafavi, S. & Wang, Y. Deep learning decodes the principles of differential gene expression. Nat. Mach. Intell. 2, 376–386 (2020).
pubmed: 32671330 pmcid: 7363043 doi: 10.1038/s42256-020-0201-6
Tao, Y. et al. Interpretable deep learning for chromatin-informed inference of transcriptional programs driven by somatic alterations across cancers. Nucleic Acids Res. 50, 10869–10881 (2022).
pubmed: 36243974 pmcid: 9638905 doi: 10.1093/nar/gkac881
Theodoris, C. V. et al. Transfer learning enables predictions in network biology. Nature 618, 616–624 (2023).
pubmed: 37258680 pmcid: 10949956 doi: 10.1038/s41586-023-06139-9
Barnett, A. J. et al. A case-based interpretable deep learning model for classification of mass lesions in digital mammography. Nat. Mach. Intell. 3, 1061–1070 (2021).
doi: 10.1038/s42256-021-00423-x
Zaritsky, A. et al. Interpretable deep learning uncovers cellular properties in label-free live cell images that are predictive of highly metastatic melanoma. Cell Syst. 12, 733–747 (2021).
pubmed: 34077708 pmcid: 8353662 doi: 10.1016/j.cels.2021.05.003
DeGrave, A. J., Cai, Z. R., Janizek, J. D., Daneshjou, R. & Lee, S. -I. Auditing the inference processes of medical-image classifiers by leveraging generative AI and the expertise of physicians. Nat. Biomed. Eng. https://doi.org/10.1038/s41551-023-01160-9 (2023).
Heil, B. J. et al. Reproducibility standards for machine learning in the life sciences. Nat. Methods 18, 1132–1135 (2021).
pubmed: 34462593 pmcid: 9131851 doi: 10.1038/s41592-021-01256-7
Whalen, S., Schreiber, J., Noble, W. S. & Pollard, K. S. Navigating the pitfalls of applying machine learning in genomics. Nat. Rev. Genet. 23, 169–181 (2022).
pubmed: 34837041 doi: 10.1038/s41576-021-00434-9
Sapoval, N. et al. Current progress and open challenges for applying deep learning across the biosciences. Nat. Commun. 13, 1728 (2022).
pubmed: 35365602 pmcid: 8976012 doi: 10.1038/s41467-022-29268-7
Chen, V., Li, J., Kim, J. S., Plumb, G. & Talwalkar, A. Interpretable machine learning: moving from mythos to diagnostics. Commun. ACM 65, 43–50 (2022). This paper describes the disconnect between IML techniques and downstream use cases and outlines paths forward to bridge the disconnect.
doi: 10.1145/3546036
Räuker, T., Ho, A., Casper, S. & Hadfield-Menell, D. Toward transparent AI: a survey on interpreting the inner structures of deep neural networks. In 2023 IEEE Conference on Secure and Trustworthy Machine Learning (SaTML), 464–483 (IEEE, 2023).
Yang, M. & Ma, J. Machine learning methods for exploring sequence determinants of 3D genome organization. J. Mol. Biol. 434, 167666 (2022).
Shrikumar, A., Greenside, P. & Kundaje, A. Learning important features through propagating activation differences. In International Conference on Machine Learning, 3145–3153 (PMLR, 2017).
Sundararajan, M., Taly, A. & Yan, Q. Axiomatic attribution for deep networks. In International Conference on Machine Learning, 3319–3328 (PMLR, 2017).
Selvaraju, R. R. et al. Grad-CAM: visual explanations from deep networks via gradient-based localization. In Proceedings of the IEEE International Conference on Computer Vision, 618–626 (2017).
Nair, S., Shrikumar, A., Schreiber, J. & Kundaje, A. fastISM: performant in silico saturation mutagenesis for convolutional neural networks. Bioinformatics 38, 2397–2403 (2022).
pubmed: 35238376 pmcid: 9048647 doi: 10.1093/bioinformatics/btac135
Lundberg, S. M. & Lee, S. -I. A unified approach to interpreting model predictions. In Proceedings of the 31st International Conference on Neural Information Processing Systems, 4768–4777 (2017).
Ribeiro, M. T., Singh, S. & Guestrin, C. “Why should I trust you?” explaining the predictions of any classifier. In Proceedings of the 22nd ACM SIGKDD, 1135–1144 (2016).
Tseng, A., Shrikumar, A. & Kundaje, A. Fourier-transform-based attribution priors improve the interpretability and stability of deep learning models for genomics. Adv. Neural Inf. Process. Syst. 33, 1913–1923 (2020).
Rudin, C. Stop explaining black box machine learning models for high stakes decisions and use interpretable models instead. Nat. Mach. Intell. 1, 206–215 (2019).
Hastie, T. & Tibshirani, R. Generalized additive models: some applications. J. Am. Stat. Assoc. 82, 371–386 (1987).
doi: 10.1080/01621459.1987.10478440
Elmarakeby, H. A. et al. Biologically informed deep neural network for prostate cancer discovery. Nature 598, 348–352 (2021).
pubmed: 34552244 pmcid: 8514339 doi: 10.1038/s41586-021-03922-4
Fortelny, N. & Bock, C. Knowledge-primed neural networks enable biologically interpretable deep learning on single-cell sequencing data. Genome Biol. 21, 190 (2020).
pubmed: 32746932 pmcid: 7397672 doi: 10.1186/s13059-020-02100-5
Janizek, J. D. et al. PAUSE: principled feature attribution for unsupervised gene expression analysis. Genome Biol. 24, 81 (2023). This paper proposes an approach to combining the post hoc and by-design explanation methods.
pubmed: 37076856 pmcid: 10114348 doi: 10.1186/s13059-023-02901-4
Vaswani, A. et al. Attention is all you need. In Advances in Neural Information Processing Systems, 5998–6008 (2017).
Karbalayghareh, A., Sahin, M. & Leslie, C. S. Chromatin interaction–aware gene regulatory modeling with graph attention networks. Genome Res. 32, 930–944 (2022).
pubmed: 35396274 pmcid: 9104700
Serrano, S. & Smith, N. A. Is attention interpretable? In Proceedings of the 57th Annual Meeting of the Association for Computational Linguistics, 2931–2951 (2019).
Jain, S. & Wallace, B. C. Attention is not explanation. In Proceedings of NAACL-HLT, 3543–3556 (2019).
Wiegreffe, S. & Pinter, Y. Attention is not not explanation. In Proceedings of EMNLP-IJCNLP, 11–20 (2019).
Bai, B. et al. Why attentions may not be interpretable? In Proceedings of the 27th ACM SIGKDD Conference on Knowledge Discovery & Data Mining, 25–34 (2021).
Conmy, A., Mavor-Parker, A. N., Lynch, A., Heimersheim, S. & Garriga-Alonso, A. Towards automated circuit discovery for mechanistic interpretability. In Thirty-seventh Conference on Neural Information Processing Systems, 16318–16352 (2023).
Friedman, D., Wettig, A. & Chen, D. Learning transformer programs. In Advances in Neural Information Processing Systems, vol. 36 (2023).
Alvarez Melis, D. & Jaakkola, T. Towards robust interpretability with self-explaining neural networks. In Advances in Neural Information Processing Systems, vol. 31 (2018).
Jacovi, A. & Goldberg, Y. Towards faithfully interpretable NLP systems: how should we define and evaluate faithfulness? In Proceedings of the 58th Annual Meeting of the Association for Computational Linguistics, 4198–4205 (ACL, 2020).
Yang, M. & Kim, B. Benchmarking attribution methods with relative feature importance. Preprint at https://arxiv.org/abs/1907.09701 (2019).
Adebayo, J., Muelly, M., Liccardi, I. & Kim, B. Debugging tests for model explanations. In Advances in Neural Information Processing Systems 33 (NIPS, 2020).
Kim, J. S., Plumb, G. & Talwalkar, A. Sanity simulations for saliency methods. In Proceedings of the 39th International Conference on Machine Learning, 11173–11200 (2022).
Zhou, Y., Booth, S., Ribeiro, M. T. & Shah, J. Do feature attribution methods correctly attribute features? In Proceedings of the AAAI Conference on Artificial Intelligence, 36, 9623–9633 (2022).
doi: 10.1609/aaai.v36i9.21196
Agarwal, C. et al. Openxai: towards a transparent evaluation of model explanations. In Advances in Neural Information Processing Systems, 35 (2022).
Ghorbani, A., Abid, A. & Zou, J. Interpretation of neural networks is fragile. In Proceedings of the AAAI Conference on Artificial Intelligence, 33, 3681–3688 (2019).
doi: 10.1609/aaai.v33i01.33013681
Krishna, S. et al. The disagreement problem in explainable machine learning: a practitioner’s perspective. Preprint at https://arxiv.org/abs/2202.01602 (2022).
Zhao, Y., Shao, J. & Asmann, Y. W. Assessment and optimization of explainable machine learning models applied to transcriptomic data. Genomics Proteomics Bioinformatics 20, 899–911 (2022).
pubmed: 35931322 pmcid: 10025763 doi: 10.1016/j.gpb.2022.07.003
Tang, Z. et al. Interpretable classification of Alzheimer’s disease pathologies with a convolutional neural network pipeline. Nat. Commun. 10, 2173 (2019).
pubmed: 31092819 pmcid: 6520374 doi: 10.1038/s41467-019-10212-1
Zeiler, M. D. & Fergus, R. Visualizing and understanding convolutional networks. In (eds. Fleet, D et al.) Computer Vision – ECCV 2014. ECCV 2014. Lecture Notes in Computer Science, vol 8689. Springer, Cham. https://doi.org/10.1007/978-3-319-10590-1_53 (2014).
Shrikumar, A. et al. Technical note on transcription factor motif discovery from importance scores (TF-MoDISco) version 0.5. 6.5. Preprint at https://arxiv.org/abs/1811.00416 (2018).
Rahman, M. A. & Rangwala, H. IDMIL: an alignment-free interpretable Deep Multiple Instance Learning (MIL) for predicting disease from whole-metagenomic data. Bioinformatics 36, i39–i47 (2020).
pubmed: 32657370 pmcid: 7355246 doi: 10.1093/bioinformatics/btaa477
Wang, L. et al. An interpretable deep-learning architecture of capsule networks for identifying cell-type gene expression programs from single-cell RNA-sequencing data. Nat. Mach. Intell. 2, 693–703 (2020).
doi: 10.1038/s42256-020-00244-4
Nagao, Y., Sakamoto, M., Chinen, T., Okada, Y. & Takao, D. Robust classification of cell cycle phase and biological feature extraction by image-based deep learning. Mol. Biol. Cell 31, 1346–1354 (2020).
pubmed: 32320349 pmcid: 7353138 doi: 10.1091/mbc.E20-03-0187
Lafarge, M. W. et al. Capturing single-cell phenotypic variation via unsupervised representation learning. In International Conference on Medical Imaging with Deep Learning, 315–325 (PMLR, 2019).
Tan, J. et al. Cell-type-specific prediction of 3D chromatin organization enables high-throughput in silico genetic screening. Nat. Biotechnol. 41, 1140–1150 (2023).
pubmed: 36624151 pmcid: 10329734 doi: 10.1038/s41587-022-01612-8
Dalla-Torre, H. et al. The nucleotide transformer: building and evaluating robust foundation models for human genomics. Preprint at bioRxiv https://doi.org/10.1101/2023.01.11.523679 (2023).
Abnar, S. & Zuidema, W. Quantifying attention flow in transformers. In Proceedings of the 58th Annual Meeting of the Association for Computational Linguistics 4190–4197 (ACL, 2020).
Tang, X. et al. Explainable multi-task learning for multi-modality biological data analysis. Nat. Commun. 14, 2546 (2023).
pubmed: 37137905 pmcid: 10156823 doi: 10.1038/s41467-023-37477-x
Washburn, J. D. et al. Evolutionarily informed deep learning methods for predicting relative transcript abundance from DNA sequence. Proc. Natl Acad. Sci. USA 116, 5542–5549 (2019).
pubmed: 30842277 pmcid: 6431157 doi: 10.1073/pnas.1814551116
Nguyen, E. et al. HyenaDNA: Long-range genomic sequence modeling at single nucleotide resolution. In Advances in Neural Information Processing Systems, 36 (2024).
Zhou, Z., Ji, Y., Li, W., Dutta, P., Davuluri, R. V. & Liu, H. DNABERT-2: efficient foundation model and benchmark for multi-species genomes. In International Conference on Learning Representations (ICLR, 2024).
Wei, J. et al. Chain-of-thought prompting elicits reasoning in large language models. Adv. Neural Inf. Process. Syst. 35, 24824–24837 (2022).
Liu, K., Casper, S., Hadfield-Menell, D. & Andreas, J. Cognitive dissonance: why do language model outputs disagree with internal representations of truthfulness? In Proceedings of the 2023 Conference on Empirical Methods in Natural Language Processing 4791–4797 (2023).
Srivastava, D., Aydin, B., Mazzoni, E. O. & Mahony, S. An interpretable bimodal neural network characterizes the sequence and preexisting chromatin predictors of induced transcription factor binding. Genome Biol. 22, 20 (2021).
Chen, R. J. et al. Pathomic fusion: an integrated framework for fusing histopathology and genomic features for cancer diagnosis and prognosis. IEEE Trans. Med. Imaging 41, 757–770(2020).
Liu, C., Huang, H. & Yang, P. Multi-task learning from multimodal single-cell omics with Matilda. Nucleic Acids Res. 51, e45 (2023).
pubmed: 36912104 pmcid: 10164589 doi: 10.1093/nar/gkad157
Liang, P. P. et al. MultiViz: towards visualizing and understanding multimodal models. In Eleventh International Conference on Learning Representations (2023).
Valeri, J. A. et al. BioAutoMATED: an end-to-end automated machine learning tool for explanation and design of biological sequences. Cell Syst. 14, 525–542 (2023).
pubmed: 37348466 pmcid: 10700034 doi: 10.1016/j.cels.2023.05.007
Ji, Y., Zhou, Z., Liu, H. & Davuluri, R. V. DNABERT: pre-trained bidirectional encoder representations from transformers model for DNA-language in genome. Bioinformatics 37, 2112–2120 (2021).
pubmed: 33538820 pmcid: 11025658 doi: 10.1093/bioinformatics/btab083

Auteurs

Valerie Chen (V)

Machine Learning Department, School of Computer Science, Carnegie Mellon University, Pittsburgh, PA, USA.

Muyu Yang (M)

Ray and Stephanie Lane Computational Biology Department, School of Computer Science, Carnegie Mellon University, Pittsburgh, PA, USA.

Wenbo Cui (W)

Machine Learning Department, School of Computer Science, Carnegie Mellon University, Pittsburgh, PA, USA.

Joon Sik Kim (JS)

Machine Learning Department, School of Computer Science, Carnegie Mellon University, Pittsburgh, PA, USA.

Ameet Talwalkar (A)

Machine Learning Department, School of Computer Science, Carnegie Mellon University, Pittsburgh, PA, USA. talwalkar@cmu.edu.

Jian Ma (J)

Ray and Stephanie Lane Computational Biology Department, School of Computer Science, Carnegie Mellon University, Pittsburgh, PA, USA. jianma@cs.cmu.edu.

Articles similaires

[Redispensing of expensive oral anticancer medicines: a practical application].

Lisanne N van Merendonk, Kübra Akgöl, Bastiaan Nuijen
1.00
Humans Antineoplastic Agents Administration, Oral Drug Costs Counterfeit Drugs

Smoking Cessation and Incident Cardiovascular Disease.

Jun Hwan Cho, Seung Yong Shin, Hoseob Kim et al.
1.00
Humans Male Smoking Cessation Cardiovascular Diseases Female
Humans United States Aged Cross-Sectional Studies Medicare Part C
1.00
Humans Yoga Low Back Pain Female Male

Classifications MeSH