{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,2,27]],"date-time":"2026-02-27T06:16:18Z","timestamp":1772172978470,"version":"3.50.1"},"update-to":[{"DOI":"10.1371\/journal.pcbi.1009880","type":"new_version","label":"New version","source":"publisher","updated":{"date-parts":[[2022,11,1]],"date-time":"2022-11-01T00:00:00Z","timestamp":1667260800000}}],"reference-count":43,"publisher":"Public Library of Science (PLoS)","issue":"10","license":[{"start":{"date-parts":[[2022,10,20]],"date-time":"2022-10-20T00:00:00Z","timestamp":1666224000000},"content-version":"vor","delay-in-days":0,"URL":"http:\/\/creativecommons.org\/licenses\/by\/4.0\/"}],"content-domain":{"domain":["www.ploscompbiol.org"],"crossmark-restriction":false},"short-container-title":["PLoS Comput Biol"],"abstract":"<jats:p>\n                    Multi-task deep learning (DL) models can accurately predict diverse genomic marks from sequence, but whether these models learn the causal relationships between genomic marks is unknown. Here, we describe Deep Mendelian Randomization (\n                    <jats:monospace>DeepMR<\/jats:monospace>\n                    ), a method for estimating causal relationships between genomic marks learned by genomic DL models. By combining Mendelian randomization with\n                    <jats:italic>in silico<\/jats:italic>\n                    mutagenesis,\n                    <jats:monospace>DeepMR<\/jats:monospace>\n                    obtains local (locus specific) and global estimates of (an assumed) linear causal relationship between marks. In a simulation designed to test recovery of pairwise causal relations between transcription factors (TFs),\n                    <jats:monospace>DeepMR<\/jats:monospace>\n                    gives accurate and unbiased estimates of the \u2018true\u2019 global causal effect, but its coverage decays in the presence of sequence-dependent confounding. We then apply\n                    <jats:monospace>DeepMR<\/jats:monospace>\n                    to examine the global relationships learned by a state-of-the-art DL model, BPNet, between TFs involved in reprogramming.\n                    <jats:monospace>DeepMR<\/jats:monospace>\n                    \u2019s causal effect estimates validate previously hypothesized relationships between TFs and suggest new relationships for future investigation.\n                  <\/jats:p>","DOI":"10.1371\/journal.pcbi.1009880","type":"journal-article","created":{"date-parts":[[2022,10,20]],"date-time":"2022-10-20T14:42:29Z","timestamp":1666276949000},"page":"e1009880","update-policy":"https:\/\/doi.org\/10.1371\/journal.pcbi.corrections_policy","source":"Crossref","is-referenced-by-count":13,"title":["Deep mendelian randomization: Investigating the causal knowledge of genomic deep learning models"],"prefix":"10.1371","volume":"18","author":[{"ORCID":"https:\/\/orcid.org\/0000-0002-7383-0094","authenticated-orcid":true,"given":"Stephen","family":"Malina","sequence":"first","affiliation":[],"role":[{"role":"author","vocabulary":"crossref"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-6908-6792","authenticated-orcid":true,"given":"Daniel","family":"Cizin","sequence":"additional","affiliation":[],"role":[{"role":"author","vocabulary":"crossref"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-7408-146X","authenticated-orcid":true,"given":"David A.","family":"Knowles","sequence":"additional","affiliation":[],"role":[{"role":"author","vocabulary":"crossref"}]}],"member":"340","published-online":{"date-parts":[[2022,10,20]]},"reference":[{"issue":"8","key":"pcbi.1009880.ref001","doi-asserted-by":"crossref","first-page":"831","DOI":"10.1038\/nbt.3300","article-title":"Predicting the sequence specificities of DNA-and RNA-binding proteins by deep learning","volume":"33","author":"B Alipanahi","year":"2015","journal-title":"Nature biotechnology"},{"issue":"10","key":"pcbi.1009880.ref002","doi-asserted-by":"crossref","first-page":"931","DOI":"10.1038\/nmeth.3547","article-title":"Predicting effects of noncoding variants with deep learning\u2013based sequence model","volume":"12","author":"J Zhou","year":"2015","journal-title":"Nature methods"},{"issue":"7","key":"pcbi.1009880.ref003","doi-asserted-by":"crossref","first-page":"990","DOI":"10.1101\/gr.200535.115","article-title":"Basset: learning the regulatory code of the accessible genome with deep convolutional neural networks","volume":"26","author":"DR Kelley","year":"2016","journal-title":"Genome research"},{"issue":"2","key":"pcbi.1009880.ref004","doi-asserted-by":"crossref","first-page":"193","DOI":"10.1186\/s12864-019-5489-4","article-title":"DeepHistone: a deep learning approach to predicting histone modifications","volume":"20","author":"Q Yin","year":"2019","journal-title":"BMC genomics"},{"issue":"1","key":"pcbi.1009880.ref005","doi-asserted-by":"crossref","first-page":"136","DOI":"10.1186\/s12859-017-1561-8","article-title":"RNA-protein binding motifs mining with a new hybrid deep learning based cross-domain knowledge integration approach","volume":"18","author":"X Pan","year":"2017","journal-title":"BMC bioinformatics"},{"key":"pcbi.1009880.ref006","first-page":"345140","article-title":"cDeepbind: A context sensitive deep learning model of RNA-protein binding","author":"S Gandhi","year":"2018","journal-title":"bioRxiv"},{"issue":"1","key":"pcbi.1009880.ref007","doi-asserted-by":"crossref","first-page":"1","DOI":"10.1038\/s41598-018-33654-x","article-title":"Deep-RBPPred: Predicting RNA binding proteins in the proteome scale based on deep learning","volume":"8","author":"J Zheng","year":"2018","journal-title":"Scientific reports"},{"issue":"3","key":"pcbi.1009880.ref008","doi-asserted-by":"crossref","first-page":"535","DOI":"10.1016\/j.cell.2018.12.015","article-title":"Predicting Splicing from Primary Sequence with Deep Learning","volume":"176","author":"K Jaganathan","year":"2019","journal-title":"Cell"},{"issue":"1","key":"pcbi.1009880.ref009","doi-asserted-by":"crossref","first-page":"94","DOI":"10.1186\/s13059-021-02273-7","article-title":"MTSplice predicts effects of genetic variants on tissue-specific splicing","volume":"22","author":"J Cheng","year":"2021","journal-title":"Genome Biol"},{"issue":"10","key":"pcbi.1009880.ref010","first-page":"1995","article-title":"Convolutional networks for images, speech, and time series","volume":"3361","author":"Y LeCun","year":"1995","journal-title":"The handbook of brain theory and neural networks"},{"issue":"4968","key":"pcbi.1009880.ref011","doi-asserted-by":"crossref","first-page":"505","DOI":"10.1126\/science.2200121","article-title":"Systematic evolution of ligands by exponential enrichment: RNA ligands to bacteriophage T4 DNA polymerase","volume":"249","author":"C Tuerk","year":"1990","journal-title":"science"},{"issue":"3","key":"pcbi.1009880.ref012","doi-asserted-by":"crossref","first-page":"354","DOI":"10.1038\/s41588-021-00782-6","article-title":"Base-resolution models of transcription-factor binding reveal soft motif syntax","volume":"53","author":"\u017d Avsec","year":"2021","journal-title":"Nature Genetics"},{"issue":"2","key":"pcbi.1009880.ref013","doi-asserted-by":"crossref","first-page":"124","DOI":"10.1038\/ng.2504","article-title":"Chromatin marks identify critical cell types for fine mapping complex trait variants","volume":"45","author":"G Trynka","year":"2013","journal-title":"Nature genetics"},{"issue":"8","key":"pcbi.1009880.ref014","doi-asserted-by":"crossref","first-page":"1133","DOI":"10.1002\/sim.3034","article-title":"Mendelian randomization: using genes as instruments for making causal inferences in epidemiology","volume":"27","author":"DA Lawlor","year":"2008","journal-title":"Statistics in medicine"},{"issue":"5830","key":"pcbi.1009880.ref015","doi-asserted-by":"crossref","first-page":"1497","DOI":"10.1126\/science.1141319","article-title":"Genome-wide mapping of in vivo protein-DNA interactions","volume":"316","author":"DS Johnson","year":"2007","journal-title":"Science"},{"issue":"4","key":"pcbi.1009880.ref016","doi-asserted-by":"crossref","first-page":"823","DOI":"10.1016\/j.cell.2007.05.009","article-title":"High-resolution profiling of histone methylations in the human genome","volume":"129","author":"A Barski","year":"2007","journal-title":"Cell"},{"issue":"8","key":"pcbi.1009880.ref017","doi-asserted-by":"crossref","first-page":"651","DOI":"10.1038\/nmeth1068","article-title":"Genome-wide profiles of STAT1 DNA association using chromatin immunoprecipitation and massively parallel sequencing","volume":"4","author":"G Robertson","year":"2007","journal-title":"Nature methods"},{"issue":"7153","key":"pcbi.1009880.ref018","doi-asserted-by":"crossref","first-page":"553","DOI":"10.1038\/nature06008","article-title":"Genome-wide maps of chromatin state in pluripotent and lineage-committed cells","volume":"448","author":"TS Mikkelsen","year":"2007","journal-title":"Nature"},{"issue":"2","key":"pcbi.1009880.ref019","doi-asserted-by":"crossref","first-page":"pdb","DOI":"10.1101\/pdb.prot5384","article-title":"DNase-seq: a high-resolution technique for mapping active gene regulatory elements across the genome from mammalian cells","volume":"2010","author":"L Song","year":"2010","journal-title":"Cold Spring Harbor Protocols"},{"issue":"1","key":"pcbi.1009880.ref020","doi-asserted-by":"crossref","first-page":"21","DOI":"10.1002\/0471142727.mb2129s109","article-title":"ATAC-seq: a method for assaying chromatin accessibility genome-wide","volume":"109","author":"JD Buenrostro","year":"2015","journal-title":"Current protocols in molecular biology"},{"key":"pcbi.1009880.ref021","unstructured":"Simonyan K, Vedaldi A, Zisserman A. Deep inside convolutional networks: Visualising image classification models and saliency maps. arXiv preprint arXiv:13126034. 2013;."},{"key":"pcbi.1009880.ref022","unstructured":"Springenberg JT, Dosovitskiy A, Brox T, Riedmiller M. Striving for simplicity: The all convolutional net. arXiv preprint arXiv:14126806. 2014;."},{"key":"pcbi.1009880.ref023","unstructured":"Shrikumar A, Greenside P, Kundaje A. Learning important features through propagating activation differences. In: Proceedings of the 34th International Conference on Machine Learning-Volume 70. JMLR. org; 2017. p. 3145\u20133153."},{"key":"pcbi.1009880.ref024","first-page":"4765","volume-title":"Advances in Neural Information Processing Systems","author":"SM Lundberg","year":"2017"},{"issue":"5","key":"pcbi.1009880.ref025","doi-asserted-by":"crossref","first-page":"e1008925","DOI":"10.1371\/journal.pcbi.1008925","article-title":"Global importance analysis: An interpretability method to quantify importance of genomic features in deep neural networks","volume":"17","author":"PK Koo","year":"2021","journal-title":"PLoS computational biology"},{"key":"pcbi.1009880.ref026","unstructured":"Lakshminarayanan B, Pritzel A, Blundell C. Simple and scalable predictive uncertainty estimation using deep ensembles. arXiv preprint arXiv:161201474. 2016;."},{"key":"pcbi.1009880.ref027","article-title":"A review of uncertainty quantification in deep learning: Techniques, applications and challenges","author":"M Abdar","year":"2021","journal-title":"Information Fusion"},{"issue":"8","key":"pcbi.1009880.ref028","doi-asserted-by":"crossref","first-page":"3770","DOI":"10.1021\/acs.jcim.0c00502","article-title":"Uncertainty quantification using neural networks for molecular property prediction","volume":"60","author":"L Hirschfeld","year":"2020","journal-title":"Journal of Chemical Information and Modeling"},{"key":"pcbi.1009880.ref029","unstructured":"Kuleshov V, Fenner N, Ermon S. Accurate uncertainties for deep learning using calibrated regression. arXiv preprint arXiv:180700263. 2018;."},{"issue":"5","key":"pcbi.1009880.ref030","doi-asserted-by":"crossref","first-page":"2333","DOI":"10.1177\/0962280215597579","article-title":"A review of instrumental variable estimators for Mendelian randomization","volume":"26","author":"S Burgess","year":"2017","journal-title":"Statistical methods in medical research"},{"issue":"3","key":"pcbi.1009880.ref031","doi-asserted-by":"crossref","first-page":"284","DOI":"10.1214\/aoms\/1177731868","article-title":"The fitting of straight lines if both variables are subject to error","volume":"11","author":"A Wald","year":"1940","journal-title":"The annals of mathematical statistics"},{"issue":"430","key":"pcbi.1009880.ref032","doi-asserted-by":"crossref","first-page":"431","DOI":"10.1080\/01621459.1995.10476535","article-title":"Two-stage least squares estimation of average causal effects in models with variable treatment intensity","volume":"90","author":"JD Angrist","year":"1995","journal-title":"Journal of the American statistical Association"},{"key":"pcbi.1009880.ref033","doi-asserted-by":"crossref","unstructured":"Zhao Q, Wang J, Hemani G, Bowden J, Small DS. Statistical inference in two-sample summary-data Mendelian randomization using robust adjusted profile score. arXiv preprint arXiv:180109652. 2018;.","DOI":"10.32614\/CRAN.package.mr.raps"},{"issue":"2","key":"pcbi.1009880.ref034","doi-asserted-by":"crossref","first-page":"512","DOI":"10.1093\/ije\/dyv080","article-title":"Mendelian randomization with invalid instruments: effect estimation and bias detection through Egger regression","volume":"44","author":"J Bowden","year":"2015","journal-title":"International journal of epidemiology"},{"issue":"4","key":"pcbi.1009880.ref035","doi-asserted-by":"crossref","first-page":"1242","DOI":"10.1093\/ije\/dyy080","article-title":"Modal-based estimation via heterogeneity-penalized weighting: model averaging for consistent and efficient estimation in Mendelian randomization when a plurality of candidate instruments are valid","volume":"47","author":"S Burgess","year":"2018","journal-title":"International journal of epidemiology"},{"issue":"R2","key":"pcbi.1009880.ref036","doi-asserted-by":"crossref","first-page":"R195","DOI":"10.1093\/hmg\/ddy163","article-title":"Evaluating the potential role of pleiotropy in Mendelian randomization studies","volume":"27","author":"G Hemani","year":"2018","journal-title":"Human molecular genetics"},{"key":"pcbi.1009880.ref037","first-page":"4697","article-title":"Bayesian deep learning and a probabilistic perspective of generalization","volume":"33","author":"AG Wilson","year":"2020","journal-title":"Advances in neural information processing systems"},{"issue":"3","key":"pcbi.1009880.ref038","first-page":"40","article-title":"meta: An R package for meta-analysis","volume":"7","author":"G Schwarzer","year":"2007","journal-title":"R news"},{"key":"pcbi.1009880.ref039","unstructured":"Finkelstein M, Shrikumar A, Kundaje A. Look at the Loss: Towards Robust Detection of False Positive Feature Interactions Learned by Neural Networks on Genomic Data. In: Proceedings of the 37th International Conference on Machine Learning (ICML 2020). The 2020 ICML Workshop on Computational Biology; 2020."},{"issue":"5","key":"pcbi.1009880.ref040","doi-asserted-by":"crossref","first-page":"2976","DOI":"10.1093\/nar\/gkt1249","article-title":"Systematic discovery and characterization of regulatory motifs in ENCODE TF binding experiments","volume":"42","author":"P Kheradpour","year":"2014","journal-title":"Nucleic acids research"},{"key":"pcbi.1009880.ref041","article-title":"fastISM: Performant in-silico saturation mutagenesis for convolutional neural networks","author":"S Nair","year":"2020","journal-title":"bioRxiv"},{"issue":"337","key":"pcbi.1009880.ref042","doi-asserted-by":"crossref","first-page":"140","DOI":"10.1080\/01621459.1972.10481216","article-title":"The isotonic regression problem and its dual","volume":"67","author":"RE Barlow","year":"1972","journal-title":"Journal of the American Statistical Association"},{"key":"pcbi.1009880.ref043","article-title":"Phenome-scale causal network discovery with bidirectional mediated Mendelian randomization","author":"BC Brown","year":"2020","journal-title":"bioRxiv"}],"updated-by":[{"DOI":"10.1371\/journal.pcbi.1009880","type":"new_version","label":"New version","source":"publisher","updated":{"date-parts":[[2022,11,1]],"date-time":"2022-11-01T00:00:00Z","timestamp":1667260800000}}],"container-title":["PLOS Computational Biology"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/dx.plos.org\/10.1371\/journal.pcbi.1009880","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2024,10,6]],"date-time":"2024-10-06T03:49:03Z","timestamp":1728186543000},"score":1,"resource":{"primary":{"URL":"https:\/\/dx.plos.org\/10.1371\/journal.pcbi.1009880"}},"subtitle":[],"editor":[{"given":"Luca","family":"Pinello","sequence":"first","affiliation":[],"role":[{"role":"editor","vocabulary":"crossref"}]}],"short-title":[],"issued":{"date-parts":[[2022,10,20]]},"references-count":43,"journal-issue":{"issue":"10","published-online":{"date-parts":[[2022,10,20]]}},"URL":"https:\/\/doi.org\/10.1371\/journal.pcbi.1009880","relation":{"has-preprint":[{"id-type":"doi","id":"10.1101\/2022.02.01.478608","asserted-by":"object"}]},"ISSN":["1553-7358"],"issn-type":[{"value":"1553-7358","type":"electronic"}],"subject":[],"published":{"date-parts":[[2022,10,20]]}}}