The DOI matches its Crossref record; 161 of the 163 references it cites were found in a registry or an index, and 1 are not checked yet.
Base-resolution models of transcription-factor binding reveal soft motif syntax
Open accessPublished version of a preprint
Abstract
The arrangement (syntax) of transcription factor (TF) binding motifs is an important part of the cis-regulatory code, yet remains elusive. We introduce a deep learning model, BPNet, that uses DNA sequence to predict base-resolution chromatin immunoprecipitation (ChIP)-nexus binding profiles of pluripotency TFs. We develop interpretation tools to learn predictive motif representations and identify soft syntax rules for cooperative TF binding interactions. Strikingly, Nanog preferentially binds with helical periodicity, and TFs often cooperate in a directional manner, which we validate using clustered regularly interspaced short palindromic repeat (CRISPR)-induced point mutations. Our model represents a powerful general approach to uncover the motifs and syntax of cis-regulatory sequences in genomics data.
Overview
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
Verification
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?
The source backs part of the claim, not all of it — or backs it for a different population, condition or comparison.
In the citing paper
“The BPNet architecture has been used widely in genomics and represents a very strong default architecture for modeling small-sized datasets from scratch through supervised learning.”
From Nucleotide Transformer: building and evaluating robust foundation models for human genomics (2025) · cites this paper as [9]
Experimental — accuracy not yet measuredThe source backs part of the claim, not all of it — or backs it for a different population, condition or comparison.
In the citing paper
“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 [7]
Experimental — accuracy not yet measuredThe source backs part of the claim, not all of it — or backs it for a different population, condition or comparison.
In the citing paper
“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 [7]
Experimental — accuracy not yet measuredThe cited source says what the claim attributes to it.
In the citing paper
“Finer resolution has potential benefits when the data support it15, but would extend the sequence length entering the quadratic complexity MHA and make the model engineering intractable on currently available hardware.”
From Effective gene expression prediction from sequence by integrating long-range interactions (2021) · cites this paper as [15]
The cited source says what the claim attributes to it.
In the citing paper
“Additionally, a deep learning algorithm applied to binding of pluripotency TFs at base resolution in mouse embryonic stem cells (ESCs) suggested a “soft motif syntax” whereby TFs help each other to bind directly to the enhancer with a lenient distance-dependency between binding sites [98].”
From Transcriptional enhancers and their communication with gene promoters (2021) · cites this paper as [98]
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
Deep Residual Learning for Image Recognition
He, K., Zhang, X., Ren, S., & Sun, J. (2016) · 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR) · doi:10.1109/cvpr.2016.90 · Cited by 235,474
Adam: A Method for Stochastic Optimization
Kingma, D. P., & Ba, J. (2014) · arXiv · doi:10.48550/arxiv.1412.6980 · Cited by 98,946
VMD: visual molecular dynamics
Humphrey, W., Dalke, A., & Schulten, K. (1996) · Journal of Molecular Graphics · doi:10.1016/0263-7855(96)00018-5 · Cited by 66,952
The Sequence Alignment/Map format and SAMtools
Li, H., Handsaker, B., Wysoker, A., Fennell, T., Ruan, J., Homer, N., Marth, G., Abecasis, G., Durbin, R., & 1000 Genome Project Data Processing Subgroup (2009) · Bioinformatics · doi:10.1093/bioinformatics/btp352 · Cited by 62,041
Fast and accurate short read alignment with Burrows-Wheeler transform
Li, H., & Durbin, R. (2009) · Bioinformatics · doi:10.1093/bioinformatics/btp324 · Cited by 58,607
Cutadapt removes adapter sequences from high-throughput sequencing reads
Martin, M. (2011) · EMBnet.journal · doi:10.14806/ej.17.1.200 · Cited by 33,889
Kimura, M. (1980) · Journal of Molecular Evolution · doi:10.1007/bf01731581 · Cited by 29,778
BEDTools: a flexible suite of utilities for comparing genomic features
Quinlan, A. R., & Hall, I. M. (2010) · Bioinformatics · doi:10.1093/bioinformatics/btq033 · Cited by 27,437
Ultrafast and memory-efficient alignment of short DNA sequences to the human genome.
Langmead, B., Trapnell, C., Pop, M., & Salzberg, S. L. (2009) · Genome Biology · doi:10.1186/gb-2009-10-3-r25 · Cited by 21,849
Hierarchical Grouping to Optimize an Objective Function
Ward, J. H. (1963) · Journal of the American Statistical Association · doi:10.1080/01621459.1963.10500845 · Cited by 21,341
Research works printed but not linked: 1
- Shrikumar, A. et al. TF-MoDISco v0.4.2.2-alpha: technical note. Preprint at arXiv https://arxiv.org/abs/1811.00416 (2018). · arXiv
Cited by
Sorted by citations
Items 401–500 of 654
Cakiroglu, S. A., Steinhauser, S., Smith, J., Wei, X., & Luscombe, N. M. (2021) · bioRxiv (Cold Spring Harbor Laboratory) · doi:10.1101/2021.03.19.436198 · Cited by 3
Designing Interpretable Convolution-Based Hybrid Networks for GenomicsPreprint
Ghotra, R., Lee, N. K., Tripathy, R. K., & Koo, P. K. (2021) · bioRxiv (Cold Spring Harbor Laboratory) · doi:10.1101/2021.07.13.452181 · Cited by 3
Neural Networks beyond explainability: Selective inference for sequence motifsPreprint
Villié, A., Veber, P., de Castro, Y., & Jacob, L. (2022) · bioRxiv (Cold Spring Harbor Laboratory) · doi:10.1101/2022.12.23.521748 · Cited by 4
“…particular, they have brought significant improvements in regulatory genomics, e.g. to predict cell-type specific transcription factor binding, gene expression, chromatin accessibility or histone modifica-tions from a DNA sequence (Zhou & Troyanskaya, 2015; Kelley et al., 2018; Avsec et al., 2021a;b).”
Motif Interactions Affect Post-Hoc Interpretability of Genomic Convolutional Neural NetworksPreprint
Lemanczyk, M., Bartoszewicz, J. M., & Renard, B. Y. (2024) · bioRxiv (Cold Spring Harbor Laboratory) · doi:10.1101/2024.02.15.580353 · Cited by 3
“Besides verifying that a model learned meaningful predictions, interpreting CNNs can lead to new insights for genomic questions [3, 4].”
BackgroundA deep profile of gene expression across 18 human cancersPreprint
Qiu, W., Dincer, A. B., Janizek, J. D., Çelik, S., Pittet, M. J., Naxerova, K., & Lee, S. (2024) · bioRxiv (Cold Spring Harbor Laboratory) · doi:10.1101/2024.03.17.585426 · Cited by 3
Satheesh, T., Singh, R. K., Gaur, P., Shahu, S., Ramakrishnan, S., Srivastava, M., Neogi, S., Choubey, S., & Ganji, M. (2025) · bioRxiv (Cold Spring Harbor Laboratory) · doi:10.1101/2025.01.06.631453 · Cited by 3
Pervasive and programmed nucleosome distortion patterns on single mammalian chromatin fibersPreprint
Yang, M. G., Richter, H. J., Wang, S., McNally, C. P., Harris, N., Dhillon, S., Maresca, M., de Wit, E., Willenbring, H., Maher, J. J., et al. (2025) · bioRxiv (Cold Spring Harbor Laboratory) · doi:10.1101/2025.01.17.633622 · Cited by 3
Wang, S. K., Li, J., Nair, S., Korasaju, R., Chen, Y., Zhang, Y., Kundaje, A. B., Liu, Y., Wang, N., & Chang, H. Y. (2025) · bioRxiv (Cold Spring Harbor Laboratory) · doi:10.1101/2025.03.21.644670 · Cited by 3
Self-citationDeep genomic models of allele-specific measurementsPreprint
Tu, X., Sasse, A., Chowdhary, K., Spiro, A. E., Yang, L., Chikina, M. D., Benoist, C. O., & Mostafavi, S. (2025) · bioRxiv (Cold Spring Harbor Laboratory) · doi:10.1101/2025.04.09.648060 · Cited by 3
“These models learn sequence features ( i.e., sequence motifs) and their combinatorial rules ( cis -regulatory grammar), which predict experimental measurements such as chromatin accessibility or gene expression in a given cellular context 8,10–12 .”
Predicting cell type-specific coverage profiles from DNA sequencePreprint
Linder, J., Yuan, H., & Kelley, D. R. (2025) · bioRxiv (Cold Spring Harbor Laboratory) · doi:10.1101/2025.06.10.658961 · Cited by 5
“…assays of human cells and tissues, as well as those of closely related species, researchers have developed accurate models for diverse regulatory processes including enhancer activity, transcription initiation, splicing, mRNA degradation, and more [3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16].”
Programming human cell type-specific gene expression via an atlas of AI-designed enhancersPreprint
Castillo-Hair, S. M., Yin, C., VandenBosch, L. S., Cherry, T. J., Meuleman, W. R. A., & Seelig, G. (2025) · bioRxiv (Cold Spring Harbor Laboratory) · doi:10.1101/2025.09.30.679565 · Cited by 7
Widespread low-affinity motifs enhance chromatin accessibility and regulatory potential in mESCsPreprint
Weilert, M., Brennan, K., Dalal, K., Krueger, S., Jiang, H., Martinez-Corral, R., & Zeitlinger, J. (2025) · bioRxiv (Cold Spring Harbor Laboratory) · doi:10.1101/2025.11.18.685822 · Cited by 3
“68 In contrast, context scores, a motif-centric expansion of in silico mutagenesis 70,71 or the necessity test, 72 measure the predicted loss when a motif is perturbed in its genomic sequence context 36 (Figure 2B, Methods).”
InfluentialSelf-citationEvolutionary transfer learning enables organism-wide inference of mammalian enhancer landscapesPreprint
Qiu, C., Daza, R. M., Welsh, I. C., Patwardhan, R. P., Martin, B. K., Li, T. W., Yang, S., Mannens, C. C. A., Winter, S. D., Kempynck, N., et al. (2026) · bioRxiv (Cold Spring Harbor Laboratory) · doi:10.62329/hxkk6249 · Cited by 4
authors:, L., Cheng, J., Taylor, K. R., Nicolaisen, L., Pan, J., Bycroft, C., Perino, M., Ward, T., Hawkes, G., Covill, L., et al. (2026) · medRxiv · doi:10.64898/2026.09.16.26363192 · Cited by 12
Self-citationGenVarLoader: an accelerated dataloader for applying deep learning to personalized genomics
Laub, D., Ho, A., Raina, L., Jaureguy, J., Klie, A., Salem, R. M., McVicker, G., & Carter, H. K. (2026) · BMC Genomics · doi:10.1186/s12864-026-13225-5 · Cited by 2
“Deep learning (DL) sequence models trained on reference genome sequences have improved our ability to predict the effects of noncoding variants on gene expression [1], transcription factor binding [2], histone modifications [1], chromatin accessibility [3], and DNA contact maps [4].”
Zhao, J., Zhao, Z., Li, H., Yu, H., Lin, H., Chen, H., Li, X., Liu, D., Wang, Y., & Wang, G. (2025) · BMC Genomics · doi:10.1186/s12864-025-11849-7 · Cited by 2
“However, unlike recent deep learning models such as Basenji [77] and BPNet [78], atSNP does not fully consider broader sequence contexts, limiting its predictive accuracy and comprehensiveness in predicting TF affinity changes.”
Wang, J., & Xu, A. (2025) · Pharmaceuticals · doi:10.3390/ph18101487 · Cited by 2
“Moreover, GelMA’s biomimetic properties such as RGD motifs facilitate cellular interactions that may synergize with epigenetic reprogramming [109–111].”
Zabihi, S., Hashemi, S., & Mansoori, E. G. (2026) · BMC Bioinformatics · doi:10.1186/s12859-026-06367-6 · Cited by 3
“For instance, BPNet [34] infers base-resolution ChIP-nexus profiles and cooperative binding syntax, while FCNsignal [35] employs fully convolutional networks to locate binding sites and motifs genome-wide.”
Machine learning for evolutionary genetics and molecular evolution
Svetec, N., Lee, U., & Zhao, L. (2026) · Trends in Genetics · doi:10.1016/j.tig.2026.01.013 · Cited by 2
Peleke, F. F., Zumkeller, S. M., Schirmer, D., Schmitz, G., Hartwig, T., Engelhorn, J., Weizel, S., Schmitt, A. O., Jores, T., & Szymański, J. (2026) · Nature Communications · doi:10.1038/s41467-026-73634-8 · Cited by 4
“While most approaches modelled DNA binding as binary classi fi cation problems, BPNet provides an approach to base-resolution protein – DNA modelling 17 .”
Evolution of CTCF binding sites in the human genome
Zhu, K., Zhuo, J., & Zhen, Y. (2026) · Molecular Biology and Evolution · doi:10.1093/molbev/msag167 · Cited by 3
Toward generalizable and interpretable AI in regulatory genomics
NAGAI, M., Murphy, A. E., Rizzo, K., & Koo, P. K. (2026) · Nature Genetics · doi:10.1038/s41588-026-02670-3 · Cited by 11
“More recent work has shifted toward quantitative profile prediction, regressing full-signal tracks across genomic windows 44,64 .”
Neural network architecture search with AMBER
Zhang, Y., Liu, Y., & Liu, X. S. (2021) · Nature Machine Intelligence · doi:10.1038/s42256-021-00350-x · Cited by 2
Automated sequence-based annotation and interpretation of the human genome
Kundaje, A. B., & Meuleman, W. R. A. (2022) · Nature Genetics · doi:10.1038/s41588-022-01123-x · Cited by 2
Self-citationDeep Learning on Chromatin Accessibility
Kim, D. S. (2023) · Methods in molecular biology · doi:10.1007/978-1-0716-2899-7_18 · Cited by 2
A fast machine learning dataloader for epigenetic tracks from BigWig files
Retel, J. S., Poehlmann, A., Chiou, J., Steffen, A., & Clevert, D.-A. (2024) · Bioinformatics · doi:10.1093/bioinformatics/btad767 · Cited by 2
“…that take a fixed length sequence found in a genomic region as an input and predict the presence of epigenetic marks measured in experiments like DNAse-seq, Chip-seq, and ATAC-seq has become popular (Zhou and Troyanskaya 2015, Kelley et al. 2016, 2018, Avsec et al. 2021, Novakovsky et al. 2023).”
W2V-repeated index: Prediction of enhancers and their strength based on repeated fragments
Xie, W., Yao, Z., Yuan, Y., Too, J., Li, F., Wang, H., Zhan, Y., Wu, X., Wang, Z., & Zhang, G. (2024) · Genomics · doi:10.1016/j.ygeno.2024.110906 · Cited by 2
Advancing Regulatory Genomics With Machine Learning
Bréhélin, L. (2024) · Bioinformatics and Biology Insights · doi:10.1177/11779322241249562 · Cited by 4
“For example, the authors of BPNet slightly change the goal of the approach 35 .”
InfluentialMainstream Artificial Intelligence Technologies in Contemporary Ophthalmology
Yin, S., Wang, L., Wu, S., Jiang, J., Qiang, W., Wang, Y., Wang, S., Shao, Y., Chen, W., & Li, Z. (2025) · Advanced Intelligent Systems · doi:10.1002/aisy.202500055 · Cited by 5
Karshenas, A., Röschinger, T., & García, H. G. (2026) · PLoS Computational Biology · doi:10.1371/journal.pcbi.1014092 · Cited by 2
“Likewise, deep convolutional neural networks, including BPNet and Enformer [39,40], have achieved high-resolution predictions of regulatory activity directly from raw DNA input.”
Finding Needles in the Haystack: Strategies for Uncovering Noncoding Regulatory Variants
Chen, Y., Paramo, M. I., Zhang, Y., Li, Y., Shah, S. R., Jin, Y., Zhang, J., Pan, X., & Yu, H. (2023) · Annual Review of Genetics · doi:10.1146/annurev-genet-030723-120717 · Cited by 3
“…convolutional neural networks (CNNs), to extract flexible DNA syntax and predict epigenomic signals such as chromatin histone modification (57, 144, 149), TF binding (4, 9, 144, 149), chromatin accessibility (57, 58, 93), transcription expression (3, 57, 148), and chromatin interaction (38, 147).”
BackgroundSpirov, A. V. (2024) · Slovo ru Baltic accent · doi:10.5922/2225-5346-2024-4-2 · Cited by 4
Wetzel, J. L., Zhang, K., & Singh, M. (2022) · bioRxiv (Cold Spring Harbor Laboratory) · doi:10.1101/2022.01.31.477772 · Cited by 2
Zarrineh, P., Darieva, Z. A., & Bobola, N. (2022) · bioRxiv (Cold Spring Harbor Laboratory) · doi:10.1101/2022.09.22.509048 · Cited by 2
“For example, users can provide motif input obtained with state-of-the-art-methods like deep learning models (Avsec et al 2021) or combine motif sets obtained with more than one method.”
MethodThe pioneer transcription factor ELF2 remodels the nucleosome near transcription start sitesPreprint
Xiao, T., Crowe‐McAuliffe, C., Ochmann, M., Du, Y., Hou, T., Chen, Y., Zhu, F., Seyednasrollah, F. S., Osmala, M., Vanharanta, S., et al. (2024) · bioRxiv (Cold Spring Harbor Laboratory) · doi:10.1101/2024.05.24.595749 · Cited by 2
“Motif discovery using BPNet ( 31 ), a deep learning model that can generate motifs based on contribution scores, resulted in identification of both a single ELF2 motif and the ELF2-HT2 motif as the most enriched double ETS motif in the peaks ( Fig.”
Decoding protein language models: insights from embedding space analysisPreprint
Rissom, P. F., Sarmiento, P. Y., Safer, J. F., Coley, C. W., Renard, B. Y., Heyne, H., & Iqbal, S. (2024) · bioRxiv (Cold Spring Harbor Laboratory) · doi:10.1101/2024.06.21.600139 · Cited by 2
Kathail, P., Shuai, R. W., Chung, R., Ye, C., Loeb, G. B., & Ioannidis, N. M. (2024) · bioRxiv (Cold Spring Harbor Laboratory) · doi:10.1101/2024.07.05.602265 · Cited by 2
Learning sequence-based regulatory dynamics in single-cell genomicsPreprint
Ibarra, I. L., Schneeberger, J., Erdogan, E., Redl, L., Martens, L. D., Klein, D., Aliee, H., & Theis, F. J. (2024) · bioRxiv (Cold Spring Harbor Laboratory) · doi:10.1101/2024.08.07.605876 · Cited by 2
“Deep learning (DL) models, such as convolutional neural networks and transformers 10,11 , improve performance and scalability.”
BackgroundKulandaisamy, A., Xia, B., Chen, H., Zhang, L., & Chen, K. (2024) · bioRxiv (Cold Spring Harbor Laboratory) · doi:10.1101/2024.08.27.609808 · Cited by 2
“Recent studies demonstrated that genetic sequence signatures hold promise to be cis-regulatory codes for some aspects of biological function, e.g., RNA structure(22–24), protein-binding(25), RNA stability(26), and promoters and enhancers activities(27, 28).”
BackgroundSun, C., Sun, Y., Xu, K., He, Z., Li, Z., Li, Y., Zong-yuan, Y., Wang, Y., Lin, X., Xu, X. S., et al. (2025) · bioRxiv (Cold Spring Harbor Laboratory) · doi:10.1101/2025.01.06.631405 · Cited by 2
WITHDRAWN: Deviation Error: assessing predictions for replicate measurements in genomics and beyondPreprint
Abdulnabi, H., & Westwood, J. T. (2025) · bioRxiv (Cold Spring Harbor Laboratory) · doi:10.1101/2025.05.29.656931 · Cited by 2
GrAdaBeam: Combining model gradients with evolutionary search for generalizable nucleic acid designPreprint
Shor, J., Strand, E., & McLean, C. Y. (2025) · Research Square · doi:10.1101/2025.06.20.660785 · Cited by 8
Liu, B. B., Shimasawa, M., Vermeulen, S. Y., Kim, S. H., Iremadze, N., Lipson, D., Shipony, Z., & Greenleaf, W. J. (2025) · bioRxiv (Cold Spring Harbor Laboratory) · doi:10.1101/2025.07.24.666684 · Cited by 2
Decoding the gene regulatory landscape through multimodal learning of protein-DNA interactionsPreprint
Tan, J., Fu, X., Ling, X., Mo, S., Bai, J., Rabadán, R., Fenyö, D., Boeke, J. D., Tsirigos, A., & Xia, B. (2025) · bioRxiv (Cold Spring Harbor Laboratory) · doi:10.1101/2025.08.17.670761 · Cited by 5
Yorzoi: Predicting RNA-seq coverage from DNA sequence in yeastPreprint
Schneider, T., Rafi, A. M., Jensen, C., Liao, D., Zhao, Y., de Boer, C. G., & Ellis, T. (2025) · bioRxiv (Cold Spring Harbor Laboratory) · doi:10.1101/2025.09.20.677345 · Cited by 5
Winter, S. D., Mannens, C. C. A., Christiaens, V., Vandepoel, R., Hu, L., Linnarsson, S., & Aerts, S. (2026) · bioRxiv (Cold Spring Harbor Laboratory) · doi:10.64898/2026.01.14.699402 · Cited by 5
Bhogale, S., Seward, C., Stubbs, L., & Sinha, S. (2023) · doi:10.1101/2023.11.09.565900 · Cited by 1
“While a number of previous studies have used deep neural networks to map sequences of segments to “functional” readouts such as epigenomic states (22–29), SEAMoD focuses specifically on prediction of differential gene expression between conditions, uses far fewer parameters and thus requires far…”
BackgroundJha, A., Hristov, B. H., Wang, X., Wang, S., Greenleaf, W. J., Kundaje, A. B., Aiden, E. L., Bertero, A., & Noble, W. S. (2026) · Nature Communications · doi:10.1038/s41467-026-72031-5 · Cited by 1
“Four ChromBPNet models predict chromatin accessibility from these four datasets using DNA sequences as input.”
InfluentialSelf-citationEpigenomic Sequencing Technologies
Chen, J., Keshavarzian, T., & Lupien, M. (2024) · Royal Society of Chemistry eBooks · doi:10.1039/9781837674916-00109 · Cited by 1
Murakami, K., Iida, K., & Okada, M. (2025) · Genes to Cells · doi:10.1111/gtc.70000 · Cited by 1
“…DNA sequences and the characteristics of cREs, such as quantitative TF binding (Avsec et al. 2021b), histone modification (Kelley et al. 2018; Avsec et al. 2021a; Yuan and Kelley 2022), and chromatin accessibility (Kelley et al. 2018; Avsec et al. 2021a; Yuan and Kelley 2022) or activity in…”
Li, Y., Shah, S. R., Ozer, A., Zhu, Y., Pan, X., Xia, T., Zhang, J., Leung, A. K.-Y., Wei, M., Lis, J. T., et al. (2026) · Nature Biotechnology · doi:10.1038/s41587-026-03218-w · Cited by 2
“2d,e), similar to a previous study 77 .”
Optimizing photosynthesis by targeting light signaling transcriptional networks
Sanyal, R., & Ranjan, A. (2025) · Journal of Experimental Botany · doi:10.1093/jxb/eraf384 · Cited by 1
IceQream: Quantitative chromosome accessibility analysis using physical TF models
Bercovich, A., Lifshitz, A., Eldar, M., Cheng, S., Ranen, R. S., Stelzer, Y., & Tanay, A. (2025) · Nature Communications · doi:10.1038/s41467-025-63925-x · Cited by 1
PRSformer: Disease Prediction from Million-Scale Individual Genotypes
Dibaeinia, P., German, C., Shringarpure, S., Auton, A., & Khan, A. A. (2025) · neural information processing systems · doi:10.52202/085713-4265 · Cited by 2
Afzal, Z., & Kumar, D. (2025) · Developmental Biology · doi:10.1016/j.ydbio.2025.11.013 · Cited by 1
Matsubara, T., Machida, S., Owusu, S. P. K., Asakura, A., Hashimoto, H., Matsuoka, M., & Nagasaki, M. (2025) · Briefings in Bioinformatics · doi:10.1093/bib/bbaf604 · Cited by 1
Peng, Y., Liu, X., Wu, J., Lin, S., Zhan, S., Li, H., Wang, J., & Zhao, X. (2025) · Briefings in Bioinformatics · doi:10.1093/bib/bbaf721 · Cited by 1
“BPNet [13] further refines this approach by predicting base-resolution ChIP-nexus binding profiles of pluripotency TFs directly from DNA sequences.”
Wang, T., & Liu, Z. (2026) · Interdisciplinary Sciences Computational Life Sciences · doi:10.1007/s12539-025-00807-2 · Cited by 1
“They are gener - ally composed of 6–12 base pairs (bp) [ 4, 5], although in some cases, the length may extend to 20 bp. Studies have shown that the binding of TFBS to TFs not only depends on the core motif but is also influenced by the surrounding sequence context [6 ].”
Pretraining improves prediction of genomic datasets across species
Huang, F., Wang, Y., Cutkosky, A., & Song, J. H. T. (2026) · Bioinformatics · doi:10.1093/bioinformatics/btag139 · Cited by 1
“…these modifications resulted in improvements over the baseline Enformer model at our smaller computational bud-get, this simplified model did not reach the 0.625 Pearson correlation coefficient reported by the Enformer authors (Avsec et al. 2021a) when trained on both human and mouse datasets.”
InfluentialReimonn, T. M., Yilmaz, V. O., Tran, H., Ng, G., Liu, D. L., & Abdennur, N. A. (2026) · Genome biology · doi:10.1186/s13059-026-04067-1 · Cited by 2
“For example, one promising avenue would be using joint PCA projections of Hi-C data as harmonized features for training emerging sequence-based deep learning models [57–62], thereby incorporating global long-range 3D genomic data into existing frameworks.”
Refining sequence-to-activity models by increasing model resolution
Chandra, N. A., Hu, Y., Buenrostro, J. D., Mostafavi, S., & Sasse, A. (2026) · Bioinformatics Advances · doi:10.1093/bioadv/vbag122 · Cited by 1
MORC3 represses a tandem repeat enhancer to regulate interferon
Krumwiede, L., Hollaus, D., Valeri, E., Schindler, K., Bazyl, M. A., Schiedlbauer, J., Becht, N. N., Jaritz, M., de Almeida, B. P., Schloissnig, S., et al. (2026) · The EMBO Journal · doi:10.1038/s44318-026-00799-9 · Cited by 1
“Additionally, the core tandem repeat units at the human MRE locus were replaced with the 7x consensus repeat sequence, and DNase hypersensitivity in indicated cell types was predicted using Enformer.”
InfluentialMao, S., Yang, C., Zhen, C., He, Z., Luo, Y., & Zhang, L. (2026) · Communications Biology · doi:10.1038/s42003-026-10462-y
“Motivated by previous studies 41 – 44 , we used a CNN to predict chromatin accessibility from DNA sequences.”
Weilert, M., Brennan, K., Dalal, K., Krueger, S., Jiang, H., Martinez-Corral, R., & Zeitlinger, J. (2026) · Cell Genomics · doi:10.1016/j.xgen.2026.101339 · Cited by 6
“67 In contrast, context scores, a motif-centric expansion of in silico mutagenesis 69,70 or the necessity test, 71 measure the predicted log-fold-change loss when a motif is mutated in its genomic sequence context 36 (Figure 2B; STAR Methods).”
InfluentialSelf-citationPredictive design of tissue-specific mammalian enhancers that function in the mouse embryo
Chen, S., Loubière, V., Hollingsworth, E. W., Ken, M., Mandlburger, N., Jacinto, S. H., Dizehchi, A., Schreiber, J. M., Kvon, E. Z., & Stark, A. (2026) · Nature Genetics · doi:10.1038/s41588-026-02729-1 · Cited by 3
“Although deep learning 2,3 has enabled enhancer prediction and design in mammalian cell lines and non-mammalian model organisms 4–10 (reviewed in a previous publication 11 ), it remains unclear whether such approaches can operate within the regulatory complexity of mammalian genomes and tissues in…”
Flexible use of conserved motifs constrains genome access in cell type evolution
Chai, C., Gibson, J., Li, P., McDonald, B. D., Pampari, A., Patel, A., Kundaje, A. B., & Wang, B. (2026) · Nature Ecology & Evolution · doi:10.1038/s41559-026-03164-5 · Cited by 8
“Recently developed deep-learning models 9,15,16 overcome this barrier by learning the sequence determinants of chromatin accessibility de novo, without requiring cross-species sequence alignment or predefined motif sets.”
InfluentialSelf-citationFan, C., Chen, K., Wang, Y., Ball, E. V., Stenson, P. D., Mort, M. E., Bacolla, A., Kehrer‐Sawatzki, H., Tainer, J. A., Cooper, D. N., et al. (2022) · Human Genetics · doi:10.1007/s00439-022-02500-6 · Cited by 1
Metadata-guided feature disentanglement for functional genomics
Rakowski, A., Monti, R., Huryn, V., Lemanczyk, M., Ohler, U., & Lippert, C. (2024) · Bioinformatics · doi:10.1093/bioinformatics/btae403 · Cited by 2
“These data have spurred the development of deep neural networks (DNNs) that predict the readouts of these experiments from DNA sequence inputs to better understand the sequence features that govern gene regulation (Zhou and Troyanskaya 2015, Kelley et al. 2016, Avsec et al. 2021b).”
Spirov, A. V. (2025) · Herald of the Russian Academy of Sciences · doi:10.1134/s101933162560088x · Cited by 1
Yi, R., Cho, K., & Bonneau, R. (2022) · Bioinformatics · doi:10.1093/bioinformatics/btac569 · Cited by 1
“A few methods have adopted the multitask learning approach in which data from multiple cell types and TFs are trained jointly in order to improve the overall model performance [16, 30, 17, 31, 18, 32, 33].”
BackgroundMethodInfluentialQuinn, T. P., Nguyen, D., Gupta, S., & Venkatesh, S. (2021) · bioRxiv (Cold Spring Harbor Laboratory) · doi:10.1101/2021.05.24.445518 · Cited by 1
Statistical methods for biological sequence analysis for DNA binding motifs and protein contacts
Roth, C. (2021) · eDiss (Georg-August-Universität Göttingen) · doi:10.53846/goediss-8835 · Cited by 1
Zuo, Z. (2022) · bioRxiv (Cold Spring Harbor Laboratory) · doi:10.1101/2022.04.19.488842 · Cited by 1
Deciphering the Impact of Genetic Variation on Human PolyadenylationPreprint
Linder, J., Kundaje, A. B., & Seelig, G. (2022) · bioRxiv (Cold Spring Harbor Laboratory) · doi:10.1101/2022.05.09.491198 · Cited by 1
“The network, which is illustrated in Figure 1C and is referred to as APARENT2, is architecturally similar to SpliceAI [27] and BPNet [37].”
MethodResultSelf-citationRomero, R., Menichelli, C., Marin, J., Lèbre, S., Lecellier, C., & Bréhélin, L. (2022) · bioRxiv (Cold Spring Harbor Laboratory) · doi:10.1101/2022.08.16.504098 · Cited by 1
“Hence, in contrast to CNN based methods [53, 4], our approach completely controls the predictive features inputted into the model.”
BackgroundMethodPredicting cell population-specific gene expression from genomic sequencePreprint
Michielsen, L., Reinders, M. J. T., & Mahfouz, A. E. (2022) · bioRxiv (Cold Spring Harbor Laboratory) · doi:10.1101/2022.09.06.506837 · Cited by 1
EUGENe: A Python toolkit for predictive analyses of regulatory sequencesPreprint
Klie, A., Stites, H., Jores, T., Solvason, J. J., Farley, E. K., & Carter, H. K. (2022) · bioRxiv (Cold Spring Harbor Laboratory) · doi:10.1101/2022.10.24.513593 · Cited by 1
Detection of new pioneer transcription factors as cell-type specific nucleosome bindersPreprint
Peng, Y., Song, W., Teif, V. B., Ovcharenko, I., Landsman, D., & Panchenko, A. R. (2023) · bioRxiv (Cold Spring Harbor Laboratory) · doi:10.1101/2023.05.10.540098 · Cited by 1
“The rapid growth of such data has further stimulated the development of computational methods to characterize cell-type specific transcription factors 24-27 .”
The evolution and mutational robustness of chromatin accessibility in DrosophilaPreprint
Khodursky, S., Zheng, E. B., Svetec, N., Durkin, S. M., Benjamin, S., Gadau, A., Wu, X., & Zhao, L. (2023) · bioRxiv (Cold Spring Harbor Laboratory) · doi:10.1101/2023.06.26.546587 · Cited by 1
Iurlaro, M., Masoni, F., Wirbelauer, C., Iskar, M., Burger, L., & Schübeler, D. (2023) · bioRxiv (Cold Spring Harbor Laboratory) · doi:10.1101/2023.07.25.550466 · Cited by 1
Detection of new pioneer transcription factors as cell-type specific nucleosome bindersPreprint
Peng, Y., Song, W., Teif, V. B., Ovcharenko, I., Landsman, D., & Panchenko, A. R. (2023) · eLife · doi:10.7554/elife.88936.1 · Cited by 1
Detection of new pioneer transcription factors as cell-type specific nucleosome bindersPreprint
Peng, Y., Song, W., Teif, V. B., Ovcharenko, I., Landsman, D., & Panchenko, A. R. (2023) · eLife · doi:10.7554/elife.88936.2 · Cited by 1
Zeng, W., Guo, H., Liu, Q., & Wong, W. H. (2024) · medRxiv · doi:10.1101/2024.10.04.24314860 · Cited by 1
“By accurately predicting the impact of genetic variants on transcription factor binding at a single base-pair resolution, BPNet enhances the precision of transcription factor binding prediction (41).”
Context-dependent epigenome rewiring during neuronal differentiationPreprint
Manelli, V., Diwakar, J., Beşkardeş, S., Gil, D. A., Forné, I., Chong, F., Imhof, A., & Bonev, B. (2024) · bioRxiv (Cold Spring Harbor Laboratory) · doi:10.1101/2024.10.18.618996 · Cited by 1
“To understand if subtle motif variations could explain the differential binding, we performed both kmer enrichment and deep learning analysis, based on BPnet and TFMoDisco 37 .”
MethodL2G: Repurposing Language Models for Genomics TasksPreprint
Cheng, W., Shen, J.-H., Khodak, M., Ma, J., & Talwalkar, A. (2024) · bioRxiv (Cold Spring Harbor Laboratory) · doi:10.1101/2024.12.09.627422 · Cited by 4
Iterative improvement of deep learning models using synthetic regulatory genomicsPreprint
Ribeiro-dos-Santos, A. M., & Maurano, M. T. (2025) · bioRxiv (Cold Spring Harbor Laboratory) · doi:10.1101/2025.02.04.636130 · Cited by 1
“Recent machine learning models including gkmSVM (Lee et al. 2015), DeepSEA (Zhou and Troyanskaya 2015), Enformer (Avsec et al. 2021a), and BPNet (Avsec et al. 2021b) have demonstrated impressive performance in predicting functional genomics data such as expression and DNA accessibility from sequence alone.”
Butts, J. C., Rong, S., Gosai, S. J., Castro, R. I., Noon, M., Adeniran, K., Ghosh, R., Sabeti, P. C., Tewhey, R., & Reilly, S. K. (2025) · bioRxiv (Cold Spring Harbor Laboratory) · doi:10.1101/2025.04.16.648420 · Cited by 3
Expanding the DNA Motif Lexicon of the Transcriptional Regulatory CodePreprint
Fan, J., Chaudhri, V. K., Bisht, D., Pease, N. A., Keshari, S., Gerges, P., Hall, D. R., Yi, V. F., Kales, S., Ho, C., et al. (2025) · bioRxiv (Cold Spring Harbor Laboratory) · doi:10.1101/2025.07.09.662874 · Cited by 2
Reimonn, T. M., Yilmaz, V. O., Tran, H., Ng, G., Liu, D. L., & Abdennur, N. A. (2025) · bioRxiv (Cold Spring Harbor Laboratory) · doi:10.1101/2025.08.07.669116 · Cited by 1
Enigma: An Efficient Model for Deciphering Regulatory GenomicsPreprint
Jung, A., Zhu, H., Li, R., Gao, A. J., Lau, T., Chu, V. S., Lim, D., Cole, C. B., Lee, L. J., Celaj, A., et al. (2025) · bioRxiv (Cold Spring Harbor Laboratory) · doi:10.64898/2025.12.18.694875 · Cited by 2
“Enigma is trained using the Poisson-multinomial loss function [46], following the approach of Borzoi and upweighting the multinomial loss term by a factor of 5.”
Predictive design of tissue-specific mammalian enhancers that function in vivo in the mouse embryoPreprint
Chen, S., Loubière, V., Hollingsworth, E. W., Jacinto, S. H., Dizehchi, A., Schreiber, J. M., Kvon, E. Z., & Stark, A. (2025) · bioRxiv (Cold Spring Harbor Laboratory) · doi:10.64898/2025.12.22.695948 · Cited by 1
DNA sequence quantitatively encodes CTCF-binding affinity at genome scalePreprint
Yin, Z., Wang, Y., Hou, G., Chen, Y., Feng, H., Zhu, J., Xie, Y., Wang, R., Zhang, Z., Wu, T., et al. (2026) · bioRxiv (Cold Spring Harbor Laboratory) · doi:10.64898/2026.01.05.696797 · Cited by 1
He, L. S., Bhatia, P., Bhuiyan, S. A., Semizoglou, E., Wang, J., Li, J., Nam, J., Luo, J. X. J., Arnhold, C., Zhu, D., et al. (2026) · bioRxiv (Cold Spring Harbor Laboratory) · doi:10.64898/2026.03.05.709931 · Cited by 1
Multi-scale dissection, compaction and derivatization of mammalian developmental enhancersPreprint
Lalanne, J., Li, T., Kajiwara, E. A. N., Huynh, C., Do, T. V., Martin, B. K., Regalado, S. G., & Shendure, J. (2026) · bioRxiv (Cold Spring Harbor Laboratory) · doi:10.64898/2026.04.20.719625 · Cited by 1
de Bellefon, S. M., Thibord, F., Auer, P. L., Blangero, J., Coban‐Akdemir, Z., Floyd, J. S., Fornage, M., Johnsen, J. M., Lange, L. A., Lewis, J. P., et al. (2020) · bioRxiv (Cold Spring Harbor Laboratory) · doi:10.1101/2020.11.23.394296
Hartonen, T., Kivioja, T., & Taipale, J. (2021) · bioRxiv (Cold Spring Harbor Laboratory) · doi:10.1101/2021.03.14.435285
Learning the mutational landscape of the cancer genomePreprint
Sherman, M. A., Yaari, A., Priebe, O., Dietlein, F., Loh, P.-R., & Berger, B. (2021) · bioRxiv · doi:10.1101/2021.08.03.454669
DeepMotifSyn: a deep learning approach to synthesize heterodimeric DNA motifs
Lin, J., Huang, L., Chen, X., Zhang, S., & Wong, K. (2021) · Briefings in Bioinformatics · doi:10.1093/bib/bbab334
Virtual Gene Concept and a Corresponding Pragmatic Research Program in Genetical Data Science
Huminiecki, Ł. (2021) · Entropy · doi:10.3390/e24010017
Metadata
- Authors
- Avsec, Ž., Weilert, M., Shrikumar, A., Krueger, S., Alexandari, A., Dalal, K., Fropf, R., McAnany, C., Gagneur, J., Kundaje, A., & Zeitlinger, J.
- Affiliations
- Google DeepMind (United Kingdom) (GB)
- Stowers Institute for Medical Research (US)
- Stanford University (US)
- Technical University of Munich (DE)
- Corresponding author
- Kundaje, Anshul
- Zeitlinger, Julia
- ORCID
- Venue
- Nature Genetics
- Published
- 2021-02-18
- Volume
- 53
- Issue
- 3
- Pages
- 354-366
- ISSN
- 1061-4036 (ISSN-L)
- 1546-1718
- PMC
- PMC8812996
- OpenAlex
- W3130819465
- Cited by
- 610
- References
- 163
- Citations by year
- 2026: 118 · 2025: 108 · 2024: 124 · 2023: 102 · 2022: 90 · 2021: 49
- FWCI
- 29.88
- Type
- journal article
- Language
- en
- Open access
- green
- doi.org
- Licence
- Locations
- Nature Genetics (publishedVersion)
- bioRxiv (Cold Spring Harbor Laboratory) (acceptedVersion, cc-by-nc)
- PubMed (publishedVersion)
- PubMed Central (acceptedVersion)
- Funders
- U.S. Department of Health & Human Services | NIH | National Human Genome Research Institute (1R01HG010211, 1U01HG009431, 1R01HG009674)
- Howard Hughes Medical Institute (International Student Research Fellowship)
- U.S. Department of Health & Human Services | NIH | National Institute of General Medical Sciences (1DP2GM123485)
- Topics
- Life Sciences › Biochemistry, Genetics and Molecular Biology › Molecular Biology › Genomics and Chromatin Dynamics
- Life Sciences › Biochemistry, Genetics and Molecular Biology › Molecular Biology › Gene Regulatory Network Analysis
- Life Sciences › Biochemistry, Genetics and Molecular Biology › Molecular Biology › Epigenetics and DNA Methylation
- Keywords
- transcription factor binding, pluripotency transcription factors, Nanog, helical periodicity, explainable deep learning
- MeSH terms
- Mouse Embryonic Stem Cells, Nanog Homeobox Protein, Deep Learning, Animals, Binding Sites, Transcription Factors, Reproducibility of Results, Neural Networks, Computer, Computational Biology, Chromatin Immunoprecipitation, Octamer Transcription Factor-3, Mice, SOXB1 Transcription Factors, Nucleotide Motifs, Clustered Regularly Interspaced Short Palindromic Repeats
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
