Sequential regulatory activity prediction across chromosomes with convolutional neural networks
Open accessPublished version of a preprint
Abstract
Models for predicting phenotypic outcomes from genotypes have important applications to understanding genomic function and improving human health. Here, we develop a machine-learning system to predict cell-type–specific epigenetic and transcriptional profiles in large mammalian genomes from DNA sequence alone. By use of convolutional neural networks, this system identifies promoters and distal regulatory elements and synthesizes their content to make effective gene expression predictions. We show that model predictions for the influence of genomic variants on gene expression align well to causal variants underlying eQTLs in human populations and can be useful for generating mechanistic hypotheses to enable fine mapping of disease loci.
The claims and statistics of this paper haven’t been checked yet.
Run the check — free, no account needed. It reads the open-access text and checks the sources it cites.
Integrity signals
Registry
Cites
Statistics
Claims the paper makes
The claims and statistics of this paper haven’t been checked yet.
Run the check — free, no account needed. It reads the open-access text and checks the sources it cites.
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?
Deep convolutional neural networks (CNNs) achieve the current state of the art at predicting gene expression from DNA sequences for the human and mouse genomes.
“Deep convolutional neural networks (CNNs) achieve the current state of the art at predicting gene expression from DNA sequences for the human and mouse genomes1–4.”
From Effective gene expression prediction from sequence by integrating long-range interactions (2021) · cites this paper as [3]
The Methods describe a deep CNN that predicts from DNA sequence, and the Results (Cell-type–specific gene expression; page unavailable) report effective gene-expression predictions, on par with models that use experimental regulatory measurements. The passages support the claim for the human data studied, but do not establish gene-expression performance in mouse or a state-of-the-art ranking across both genomes. (LLM, conf=0.86)
Experimental — accuracy not yet measuredMachine-learning models trained to predict function from DNA sequences have been successful at characterizing regulatory syntax and interpreting genetic variant effects.
“Machine-learning models trained to predict function from DNA sequences have been successful at characterizing regulatory syntax and interpreting genetic variant effects.”
From Predicting RNA-seq coverage from DNA sequence as a unifying model of gene regulation (2025) · cites this paper as [8]
Thus far, sequence-based models have focused on assays in which measured activity is proportional to local sequencing read counts.
“Thus far, such models have focused on assays in which measured activity is proportional to local sequencing read counts.”
From Predicting RNA-seq coverage from DNA sequence as a unifying model of gene regulation (2025) · cites this paper as [8]
Including experimental assays studied by Enformer and Basenji in training data helps the model use important sequence features for making RNA coverage predictions.
“To help the model use important sequence features for making its RNA coverage predictions, we also included the experimental assays studied by the Enformer and Basenji models in the training data8,9,13.”
From Predicting RNA-seq coverage from DNA sequence as a unifying model of gene regulation (2025) · cites this paper as [8]
This performance increase is twice as large as the performance increase between Basenji1 and Basenji2 and closes one-third of the gap to experimental-level accuracy, estimated at 0.94.
“This performance increase is twice as large as the performance increase between Basenji1 (ref. 3) and Basenji2 (ref. 2) and closes one-third of the gap to experimental-level accuracy, estimated at 0.94 (Extended Data Fig. 2).”
From Effective gene expression prediction from sequence by integrating long-range interactions (2021) · cites this paper as [3]
Self-check
Percentages, sums and reported differences re-computed from the paper’s own numbers.
The self-check needs the paper’s full text.
The citation lists haven’t been fetched yet
References 69 · Cited by 537
Research works printed but not linked: 51
- Abadi M, Barham P, Chen J, Chen Z, Davis A, Dean J, Devin M, Ghemawat S, Irving G, Isard M, 2016 TensorFlow: a system for large-scale machine learning. In USENIX symposium on operating systems design and implementation, Savannah, GA.
- Alipanahi B, Delong A, Weirauch MT, Frey BJ. 2015 Predicting the sequence specificities of DNA- and RNA-binding proteins by deep learning. Nat Biotechnol 33: 831– 838. 26213851
- Ayoubi TA, Van De Ven WJ. 1996 Regulation of gene expression by alternative promoters. FASEB J 10: 453– 460. 8647344
- Bach S, Binder A, Montavon G, Klauschen F, Müller K-R, Samek W. 2015 On pixel-wise explanations for non-linear classifier decisions by layer-wise relevance propagation. PLoS One 10: e0130140. 26161953 PMC4498753
- Benjamini Y, Speed TP. 2012 Summarizing and correcting the GC content bias in high-throughput sequencing. Nucleic Acids Res 40: e72. 22323520 PMC3378858
- Boyle EA, Li YI, Pritchard JK. 2017 An expanded view of complex traits: from polygenic to omnigenic. Cell 169: 1177– 1186. 28622505 PMC5536862
- Brinkmann V, Billich A, Baumruker T, Heining P, Schmouder R, Francis G, Aradhye S, Burtin P. 2010 Fingolimod (FTY720): discovery and development of an oral drug to treat multiple sclerosis. Nat Rev Drug Discov 9: 883– 897. 21031003
- Cheng C, Alexander R, Min R, Leng J, Yip KY, Rozowsky J, Yan K-K, Dong X, Djebali S, Ruan Y, 2012 Understanding transcriptional regulation by integrative analysis of transcription factor binding data. Genome Res 22: 1658– 1667. 22955978 PMC3431483
- Claussnitzer M, Dankel SN, Kim K-H, Quon G, Meuleman W, Haugen C, Glunk V, Sousa IS, Beaudry JL, Puviindran V, 2015 FTO obesity variant circuitry and adipocyte browning in humans. N Engl J Med 373: 895– 907. 26287746 PMC4959911
- Cusanovich DA, Pavlovic B, Pritchard JK, Gilad Y. 2014 The functional consequences of variation in transcription factor binding. PLoS Genet 10: e1004226. 24603674 PMC3945204
- Dekker J, Mirny LA. 2016 The 3D genome as moderator of chromosomal communication. Cell 164: 1110– 1121. 26967279 PMC4788811
- The ENCODE Project Consortium. 2012 An integrated encyclopedia of DNA elements in the human genome. Nature 489: 57– 74. 22955616 PMC3439153
- Ernst J, Kellis M. 2010 Discovery and characterization of chromatin states for systematic annotation of the human genome. Nat Biotechnol 28: 817– 825. 20657582 PMC2919626
- Farh KK-H, Marson A, Zhu J, Kleinewietfeld M, Housley WJ, Beik S, Shoresh N, Whitton H, Ryan RJH, Shishkin AA, 2015 Genetic and epigenetic fine mapping of causal autoimmune disease variants. Nature 518: 337– 343. 25363779 PMC4336207
- Feschotte C. 2008 Transposable elements and the evolution of regulatory networks. Nat Rev Genet 9: 397– 405. 18368054 PMC2596197
- Finucane HK, Bulik-Sullivan B, Gusev A, Trynka G, Reshef Y, Loh P-R, Anttila V, Xu H, Zang C, Farh K, 2015 Partitioning heritability by functional annotation using genome-wide association summary statistics. Nat Genet 47: 1228– 1235. 26414678 PMC4626285
- Forrest ARR, Kawaji H, Rehli M, Kenneth Baillie J, de Hoon MJL, Haberle V, Lassmann T, Kulakovskiy IV, Lizio M, Itoh M, 2014 A promoter-level mammalian expression atlas. Nature 507: 462– 470. 24670764 PMC4529748
- Ghandi M, Lee D, Mohammad-Noori M, Beer MA. 2014 Enhanced regulatory sequence prediction using gapped k-mer features. PLoS Comput Biol 10: e1003711. 25033408 PMC4102394
- Glorot X, Bengio Y. 2010 Understanding the difficulty of training deep feedforward neural networks. In Proceedings of the 13th international conference on artificial intelligence and statistics (AISTATS), Sardinia, Italy.
- González AJ, Setty M, Leslie CS. 2015 Early enhancer establishment and regulatory locus complexity shape transcriptional programs in hematopoietic differentiation. Nat Genet 47: 1249– 1259. 26390058 PMC4626279
- The GTEx Consortium. 2017 Genetic effects on gene expression across human tissues. Nature 550: 204– 213. 29022597 PMC5776756
- Gupta S, Stamatoyannopoulos JA, Bailey TL, Noble WS. 2007 Quantifying similarity between motifs. Genome Biol 8: R24. 17324271 PMC1852410
- Hashimoto T, Sherwood RI, Kang DD, Rajagopal N, Barkal AA, Zeng H, Emons BJM, Srinivasan S, Jaakkola T, Gifford DK. 2016 A synergistic DNA logic predicts genome-wide chromatin accessibility. Genome Res 10: 1430– 1440. PMC5052050 27456004
- International Multiple Sclerosis Genetics Consortium. 2013 Analysis of immune-related loci identifies 48 new susceptibility variants for multiple sclerosis. Nat Genet 45: 1353– 1360. 24076602 PMC3832895
- Kang Z, Grauman K, Sha F. 2011 Learning with whom to share in multi-task feature learning. In International conference on machine learning, Bellevue, WA.
- Kelley DR, Snoek J, Rinn JL. 2016 Basset: learning the regulatory code of the accessible genome with deep convolutional neural networks. Genome Res 26: 990– 999. 27197224 PMC4937568
- Kingma DP, Ba J. 2015 Adam: a method for stochastic optimization. In International conference on learning representations, San Diego, CA.
- Koscielny G, An P, Carvalho-Silva D, Cham JA, Fumis L, Gasparyan R, Hasan S, Karamanis N, Maguire M, Papa E, 2017 Open Targets: a platform for therapeutic target identification and validation. Nucleic Acids Res 45: D985– D994. 27899665 PMC5210543
- Langmead B, Salzberg SL. 2012 Fast gapped-read alignment with Bowtie 2. Nat Methods 9: 357– 359. 22388286 PMC3322381
- Lee D, Gorkin DU, Baker M, Strober BJ, Asoni AL, McCallion AS, Beer MA. 2015 A method to predict the impact of regulatory variants from DNA sequence. Nat Genet 47: 955– 961. 26075791 PMC4520745
- Lenhard B, Sandelin A, Carninci P. 2012 Metazoan promoters: emerging characteristics and insights into transcriptional regulation. Nat Rev Genet 13: 233– 245. 22392219
- Levine M. 2010 Transcriptional enhancers in animal development and evolution. Curr Biol 20: R754– R763. 20833320 PMC4280268
- Li YI, van de Geijn B, Raj A, Knowles DA, Petti AA, Golan D, Gilad Y, Pritchard JK. 2016 RNA splicing is a primary link between genetic variation and disease. Science 352: 600– 604. 27126046 PMC5182069
- Liu X, Finucane HK, Gusev A, Bhatia G, Gazal S, O'Connor L, Bulik-Sullivan B, Wright FA, Sullivan PF, Neale BM, 2017 Functional architectures of local and distal regulation of gene expression in multiple human tissues. Am J Hum Genet 100: 605– 616. 28343628 PMC5384099
- Long HK, Prescott SL, Wysocka J. 2016 Ever-changing landscapes: transcriptional enhancers in development and evolution. Cell 167: 1170– 1187. 27863239 PMC5123704
- MacArthur J, Bowler E, Cerezo M, Gil L, Hall P, Hastings E, Junkins H, McMahon A, Milano A, Morales J, 2017 The new NHGRI-EBI Catalog of published genome-wide association studies (GWAS Catalog). Nucleic Acids Res 45: D896– D901. 27899670 PMC5210590
- Mathelier A, Fornes O, Arenillas DJ, Chen C-Y, Denay G, Lee J, Shi W, Shyr C, Tan G, Worsley-Hunt R, 2016 JASPAR 2016: a major expansion and update of the open-access database of transcription factor binding profiles. Nucleic Acids Res 44: D110– D115. 26531826 PMC4702842
- Meyer CA, Liu XS. 2014 Identifying and mitigating bias in next-generation sequencing methods for chromatin biology. Nat Rev Genet 15: 709– 721. 25223782 PMC4473780
- Mifsud B, Tavares-Cadete F, Young AN, Sugar R, Schoenfelder S, Ferreira L, Wingett SW, Andrews S, Grey W, Ewels PA, 2015 Mapping long-range promoter contacts in human cells with high-resolution capture Hi-C. Nat Genet 47: 598– 606. 25938943
- O'Connor LJ, Gusev A, Liu X, Loh P-R, Finucane HK, Price AL. 2017 Estimating the proportion of disease heritability mediated by gene expression levels. bioRxiv 10.1101/118018. · DOI
- Reshef YA, Finucane HK, Kelley DR, Gusev A, Kotliar D, Ulirsch JC, Hormozdiari F, O'Connor L, van de Geijn B, Loh P-R, 2017 Detecting genome-wide directional effects of transcription factor binding on polygenic disease risk. bioRxiv 10.1101/204685. PMC6202062 30177862 · DOI
- Shrikumar A, Greenside P, Kundaje A. 2017 Learning important features through propagating activation differences. In International conference on machine learning, Sydney, Australia.
- Skalska L, Beltran-Nebot M, Ule J, Jenner RG. 2017 Regulatory feedback from nascent RNA to chromatin and transcription. Nat Rev Mol Cell Biol 18: 331– 337. 28270684
- Skene PJ, Henikoff S. 2017 An efficient targeted nuclease strategy for high-resolution mapping of DNA binding sites. eLife 6: e21856. 28079019 PMC5310842
- Teng M, Irizarry RA. 2016 Accounting for GC-content bias reduces systematic errors and batch effects in ChIP-Seq data. Genome Res 11: 1930– 1938. PMC5668949 29025895
- Wang J-Q, Kon J, Mogi C, Tobo M, Damirin A, Sato K, Komachi M, Malchinkhuu E, Murata N, Kimura T, 2004 TDAG8 is a proton-sensing and psychosine-sensitive G-protein-coupled receptor. J Biol Chem 279: 45626– 45633. 15326175
- Whitaker JW, Chen Z, Wang W. 2015 Predicting the human epigenome from DNA motifs. Nat Methods 12: 265– 272. 25240437 PMC4344378
- Yu F, Koltun V. 2016 Multi-scale context aggregation by dilated convolutions. In International conference on learning representations, San Juan, Puerto Rico.
- Zhang Q, Keles¸ S. 2014 CNV-guided multi-read allocation for ChIP-seq. Bioinformatics 30: 2860– 2867. 24966364 PMC4184254
- Zhang Y, Liu T, Meyer CA, Eeckhoute J, Johnson DS, Bernstein BE, Nusbaum C, Myers RM, Brown M, Li W, 2008 Model-based Analysis of ChIP-Seq (MACS). Genome Biol 9: R137. 18798982 PMC2592715
- Zhou J, Troyanskaya OG. 2015 Predicting effects of noncoding variants with deep learning–based sequence model. Nat Methods 12: 931– 934. 26301843 PMC4768299
- Authors
- Kelley, D. R., Reshef, Y. A., Bileschi, M., Belanger, D., McLean, C. Y., & Snoek, J.
- Affiliations
- Calico Life Sciences (US)
- Harvard University (US)
- Google (United States) (US)
- Venue
- Genome Research
- Publisher
- Cold Spring Harbor Laboratory
- Published
- 2018-03-27
- Volume
- 28
- Issue
- 5
- Pages
- 739-750
- ISSN
- 1088-9051 (ISSN-L)
- 1549-5469
- PMC
- PMC5932613
- OpenAlex
- W2952239877
- Cited by
- 537
- References
- 69
- Citations by year
- 2026: 86 · 2025: 68 · 2024: 87 · 2023: 82 · 2022: 75 · 2021: 53 · 2020: 48 · 2019: 24 · 2018: 13
- Type
- posted content
- Language
- en
- Open access
- green
- doi.org
- Licence
- cc-by-nc-nd
- Locations
- Genome Research (publishedVersion)
- bioRxiv (Cold Spring Harbor Laboratory) (acceptedVersion, cc-by-nc-nd)
- Europe PMC (PubMed Central) (publishedVersion, cc-by)
- PubMed Central (publishedVersion, cc-by)
- PubMed (publishedVersion)
- HighWire Press Open Archive (submittedVersion)
- Zenodo (CERN European Organization for Nuclear Research) (submittedVersion, other-oa)
- Funders
- National Institute of General Medical Sciences (T32GM007753)
- Topics
- Life Sciences › Biochemistry, Genetics and Molecular Biology › Molecular Biology › Genomics and Chromatin Dynamics
- Life Sciences › Biochemistry, Genetics and Molecular Biology › Genetics › Genetic Associations and Epidemiology
- Life Sciences › Biochemistry, Genetics and Molecular Biology › Molecular Biology › Epigenetics and DNA Methylation
- Keywords
- regulatory element prediction, promoter, distal regulatory elements, gene expression prediction, expression quantitative trait loci, causal variants, gene expression profiling, DNA sequences, convolutional neural networks, mammalian genomes
- MeSH terms
- Machine Learning, Animals, Chromosomes, Gene Expression Regulation, Humans, Models, Genetic, Promoter Regions, Genetic, Regulatory Sequences, Nucleic Acid, Neural Networks, Computer, Computational Biology, Polymorphism, Single Nucleotide, Gene Expression Profiling, Genomics, Epigenomics
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
