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 201–300 of 654
Active learning of enhancers and silencers in the developing neural retina
Friedman, R. Z., Ramu, A., Lichtarge, S., Wu, Y., Tripp, L., Lyon, D., Myers, C. A., Granas, D. M., Gause, M., Corbo, J. C., et al. (2025) · Cell Systems · doi:10.1016/j.cels.2024.12.004 · Cited by 22
Machine-guided design of synthetic cell type-specific cis -regulatory elementsPreprint
Gosai, S. J., Castro, R. I., Fuentes, N., Butts, J. C., Kales, S., Noche, R. R., Mouri, K., Sabeti, P. C., Reilly, S. K., & Tewhey, R. (2023) · bioRxiv (Cold Spring Harbor Laboratory) · doi:10.1101/2023.08.08.552077 · Cited by 18
He, A. Y., & Danko, C. G. (2024) · bioRxiv (Cold Spring Harbor Laboratory) · doi:10.1101/2024.03.13.583868 · Cited by 24
“CLIPNET’s architecture incorporates recent advances in predicting genome-wide molecular assays at single nucleotide resolution, most notably those utilized in BPNet [38] and APARENT 1 [44] and 2 [45].”
BackgroundMethodInfluentialDistinctive DNA sequence features define epigenetic longevity of inflammatory memory
Cowley, C. J., Sajjath, S. M., Soto-Ugaldi, L. F., Steiger, M., Larsen, S. B., Carroll, T., Barrows, D., Mattei, A. L., Gonzales, K. A. U., Wang, W., et al. (2026) · Science · doi:10.1126/science.adz6830 · Cited by 19
Deep learning identifies and quantifies recombination hotspot determinants
Li, Y., Chen, S., Rapakoulia, T., Kuwahara, H., Yip, K. Y.-L., & Gao, X. (2022) · Bioinformatics · doi:10.1093/bioinformatics/btac234 · Cited by 15
“Inspired by DeepLIFT, a more advanced TF-MoDISco(48) introduced for 480 transcription factor prediction was proposed.”
InfluentialToward a comprehensive catalog of regulatory elements
Fan, K., Pfister, E. L., & Weng, Z. (2023) · Human Genetics · doi:10.1007/s00439-023-02519-3 · Cited by 19
Prediction of cooperative homeodomain DNA binding sites from high-throughput-SELEX data
Cain, B., Webb, J. A., Yuan, Z., Cheung, D., Lim, H., Kovall, R. A., Weirauch, M. T., & Gebelein, B. A. (2023) · Nucleic Acids Research · doi:10.1093/nar/gkad318 · Cited by 15
Genomic context sensitizes regulatory elements to genetic disruption
Ordóñez, R., Zhang, W., Ellis, G., Zhu, Y., Ashe, H. J., Ribeiro-dos-Santos, A. M., Brosh, R., Huang, E., Hogan, M. S., Boeke, J. D., et al. (2024) · Molecular Cell · doi:10.1016/j.molcel.2024.04.013 · Cited by 22
Genetic coupling of enhancer activity and connectivity in gene expression control
Ray-Jones, H. F., Sung, C. K., Chan, L. T., Haglund, A., Artemov, P., Rosa, M. D., Ruje, L., Burden, F., Kreuzhuber, R., Litovskikh, A., et al. (2025) · Nature Communications · doi:10.1038/s41467-025-55900-3 · Cited by 17
Deep exploration networks for rapid engineering of functional DNA sequencesPreprint
Linder, J., Bogard, N., Rosenberg, A. B., & Seelig, G. (2019) · bioRxiv (Cold Spring Harbor Laboratory) · doi:10.1101/864363 · Cited by 17
Machine learning for profile prediction in genomics
Schreiber, J. M., & Singh, R. (2021) · Current Opinion in Chemical Biology · doi:10.1016/j.cbpa.2021.04.008 · Cited by 15
Schreiber, J. M., Boix, C. A., Lee, J. W., Li, H., Guan, Y., Chang, C.-C., Chang, J.-C., Hawkins‐Hooker, A., Schölkopf, B., Schweikert, G., et al. (2023) · Genome biology · doi:10.1186/s13059-023-02915-y · Cited by 16
“Potentially, future iterations of the imputation work could involve imputing read counts but allowing models to directly incorporate the control tracks and other covariates such as sequencing depth, single-end or paired-end status, and data quality metrics as well [18].”
BackgroundSelf-citationBerenson, A., Lane, R., Soto-Ugaldi, L. F., Patel, M., Ciausu, C., Li, Z., Chen, Y., Shah, S., Santoso, C., Liu, X., et al. (2023) · Nature Communications · doi:10.1038/s41467-023-42445-6 · Cited by 15
Modeling and designing enhancers by introducing and harnessing transcription factor binding units
Li, J., Zhang, P., Xi, X., Liu, L., Wei, L., & Wang, X. (2025) · Nature Communications · doi:10.1038/s41467-025-56749-2 · Cited by 16
Discovering differential genome sequence activity with interpretable and efficient deep learning
Hammelman, J., & Gifford, D. (2021) · bioRxiv · doi:10.1371/journal.pcbi.1009282 · Cited by 19
“ChIP-nexus binding data for Oct4, Sox2, and patch cap control [17] were downloaded as fastqs from the Nucleotide Read Archive (see S1 Table for accession numbers).”
MethodInterplay between regulatory elements and chromatin topology in cellular lineage determination
Shukla, V., Cetnarowska, A., Hyldahl, M., & Mandrup, S. (2022) · Trends in Genetics · doi:10.1016/j.tig.2022.05.011 · Cited by 15
BindVAE: Dirichlet variational autoencoders for de novo motif discovery from accessible chromatin
Kshirsagar, M., Yuan, H., Ferres, J. L., & Leslie, C. S. (2022) · Genome biology · doi:10.1186/s13059-022-02723-w · Cited by 16
“Interpretation of supervised deep sequence models has been more successful where models are trained on high-resolution TF occupancy data [12], since the underlying motif grammar is less complex than that for chromatin accessibility data.”
BackgroundMoyers, B. A., Partridge, E. C., Mackiewicz, M., Betti, M. J., Darji, R., Meadows, S. K., Newberry, K. M., Brandsmeier, L. A., Wold, B. J., Mendenhall, E. M., et al. (2023) · Genome Research · doi:10.1101/gr.278205.123 · Cited by 16
“Promising tools are being developed to predict the activity of factors (Schreiber et al. 2020; Avsec et al. 2021), yet these tools require high-quality data sets in defined contexts for accurate training.”
BackgroundMultiomics and deep learning dissect regulatory syntax in human development
Liu, B. B., Jessa, S., Kim, S. H., Ng, Y. T., Higashino, S. I., Marinov, G. K., Chen, D. C., Parks, B. E., Li, L., Nguyen, T. C., et al. (2026) · Nature · doi:10.1038/s41586-026-10326-9 · Cited by 36
Self-citationTangermeme: a toolkit for understanding cis-regulatory logic using deep learning models
Schreiber, J. M. (2026) · Nature Methods · doi:10.1038/s41592-026-03254-z · Cited by 16
Deep mendelian randomization: Investigating the causal knowledge of genomic deep learning models
Malina, S., Cizin, D. M., & Knowles, D. A. (2022) · PLoS Computational Biology · doi:10.1371/journal.pcbi.1009880 · Cited by 15
“We used the ChIP-nexus data and model (BPNet) previously described in [Avsec et al., 2020] but trained a 5-component ensemble.”
MethodDeep learning approaches for noncoding variant prioritization in neurodegenerative diseases
Lan, A. Y., & Corces, M. R. (2022) · Frontiers in Aging Neuroscience · doi:10.3389/fnagi.2022.1027224 · Cited by 16
“…multitude of biological phenomena and data types including but not limited to general chromatin features (Kelley et al., 2018), TF binding profiles (Avsec et al., 2021b), gene expression (Zhou et al., 2018), 3D chromatin conformation (Fudenberg et al., 2020), DNA methylation (Angermueller et al.,…”
BackgroundDuncan, A. G., Mitchell, J. A., & Moses, A. M. (2024) · Bioinformatics · doi:10.1093/bioinformatics/btae190 · Cited by 16
“…et al. 2020, Koo et al. 2021) has revealed that deep learning models trained on regulatory sequences learn both known and novel motifs and motif interactions (Kelley et al. 2016, Maslova et al. 2020, Minnoye et al. 2020, Avsec et al. 2021a,b, de Almeida et al. 2022, Novakovsky et al. 2023).”
BackgroundCross-species analysis of melanoma enhancer logic using deep learningPreprint
Minnoye, L., Taskiran, I. I., Mauduit, D., Fazio, M., Aerschot, L. V., Hulsemans, G., Christiaens, V., Makhzami, S., Seltenhammer, M. H., Karras, P., et al. (2019) · bioRxiv (Cold Spring Harbor Laboratory) · doi:10.1101/2019.12.21.885715 · Cited by 15
Prakash, E., Shrikumar, A., & Kundaje, A. (2021) · bioRxiv · doi:10.1101/2021.12.26.474224 · Cited by 19
Self-citationMa, X. R., Conley, S. D., Chen, M. S., Kosicki, M., Lu, Y., Bredikhin, D., Cui, R., Sheth, M. U., Qiu, W., Amgalan, D., et al. (2024) · medRxiv · doi:10.1101/2024.11.20.24317557 · Cited by 18
Self-citationDecoding polygenic diseases: advances in noncoding variant prioritization and validation
Chin, I. M., Gardell, Z. A., & Corces, M. R. (2024) · Trends in Cell Biology · doi:10.1016/j.tcb.2024.03.005 · Cited by 18
A multi-modal transformer for cell type-agnostic regulatory predictions
Javed, N. M., Weingarten, T., Sehanobish, A., Roberts, A., Dubey, A., Choromański, K., & Bernstein, B. (2025) · Cell Genomics · doi:10.1016/j.xgen.2025.100762 · Cited by 22
Uncertainty-aware genomic deep learning with knowledge distillation
Zhou, J., Rizzo, K., Christensen, T. A., Tang, Z., & Koo, P. K. (2026) · npj Artificial Intelligence · doi:10.1038/s44387-025-00053-3 · Cited by 22
Detection of new pioneer transcription factors as cell-type-specific nucleosome binders
Peng, Y., Song, W., Teif, V. B., Ovcharenko, I., Landsman, D., & Panchenko, A. R. (2023) · eLife · doi:10.7554/elife.88936 · Cited by 22
“The rapid growth of such datasets has further stimulated the development of computational methods to characterize cell- type-specific TFs ( Sherwood et al., 2014 ; Jankowski et al., 2016 ; Zheng et al., 2021 ; Avsec et al., 2021 ).”
BackgroundMethodZhang, Q., Wang, S., Li, Z., Pan, Y., & Huang, D. (2024) · Advanced Science · doi:10.1002/advs.202405685 · Cited by 15
MorPhiC Consortium: towards functional characterization of all human genes
Adli, M., Przybyla, L., Burdett, T., Burridge, P. W., Cacheiro, P., Chang, H. Y., Engreitz, J., Gilbert, L. A., Greenleaf, W. J., Hsu, L.-I., et al. (2025) · Nature · doi:10.1038/s41586-024-08243-w · Cited by 21
“…to infer how genes act together in regulatory networks and pathways—including, for example, approaches leveraging inference of gene-expression programs 33,81–83 , gene regulatory networks linking transcription factors to target genes 47–49, 84, 85 , and RNA velocity and trajectory analyses”
Self-citationPerspective on recent developments and challenges in regulatory and systems genomics
Zeitlinger, J., Roy, S., Ay, F., Mathelier, A., Medina-Rivera, A., Mahony, S., Sinha, S. K., & Ernst, J. (2024) · Bioinformatics Advances · doi:10.1093/bioadv/vbaf106 · Cited by 18
“This is possible and yields highly accurate predictions (Kelley et al. 2018, Avsec et al. 2021a, Linder et al. 2025) and promoter syntax (Dudnyk et al. 2024, Cochran et al. 2024, He and Danko 2024).”
InfluentialSelf-citationAccelerating in silico saturation mutagenesis using compressed sensing
Schreiber, J. M., Nair, S., Balsubramani, A., & Kundaje, A. B. (2022) · Bioinformatics · doi:10.1093/bioinformatics/btac385 · Cited by 16
Self-citationKarollus, A., Mauermeier, T., & Gagneur, J. (2022) · bioRxiv (Cold Spring Harbor Laboratory) · doi:10.1101/2022.09.15.508087 · Cited by 14
Self-citationChatNT: A Multimodal Conversational Agent for DNA, RNA and Protein TasksPreprint
Richard, G., de Almeida, B. P., Dalla-Torre, H., Blum, C. F., Hexemer, L., Pandey, P., Laurent, S., Lopez, M., Laterre, A., Lang, M., et al. (2024) · bioRxiv (Cold Spring Harbor Laboratory) · doi:10.1101/2024.04.30.591835 · Cited by 14
Badia-i-Mompel, P., Casals-Franch, R., Wessels, L., Müller‐Dott, S., Trimbour, R., Yang, Y., Flores, R. O. R., & Sáez-Rodríguez, J. (2024) · bioRxiv (Cold Spring Harbor Laboratory) · doi:10.1101/2024.12.20.629764 · Cited by 16
Best practices for interpretable machine learning in computational biology
Chen, V., Yang, M., Cui, W., Kim, J. S., Talwalkar, A., & Ma, J. (2022) · doi:10.1101/2022.10.28.513978 · Cited by 14
“To showcase explanation verification, we utilize the example of BPNet (Avsec et al., 2021), which is trained on high-resolution ChIP–nexus data to predict base-resolution TF binding, and the same IML method, DeepLIFT (Shrikumar et al., 2017), to interpret the BPNet. DeepLIFT is used to identify transcription factor bi…”
BackgroundMethodInfluentialLally, P., Gómez-Romero, L., Tierrafría, V. H., Aquino, P., Rioualen, C., Zhang, X., Kim, S., Baniulyte, G., Plitnick, J., Smith, C., et al. (2025) · Nature Communications · doi:10.1038/s41467-025-58862-8 · Cited by 13
“BPNet is a deep convolutional neural network designed to predict genome coverage from sequence 13 .”
Włodarczyk, T., Lun, A. T. L., Wu, D. H., Shi, M., Ye, X., Menon, S., Toneyan, S., Seidel, K., Wang, L., Tan, J., et al. (2025) · Nature Communications · doi:10.1038/s41467-025-62252-5 · Cited by 18
Li, X., Ye, C., Liu, T., Li, S., Zhang, M., Zhao, Y., Jin, Y., Cheng, J., Yang, G., & Li, P. (2025) · Synthetic and Systems Biotechnology · doi:10.1016/j.synbio.2025.10.008 · Cited by 17
Regulatory grammar in human promoters uncovered by MPRA-based deep learning
Barbadilla-Martínez, L., Klaassen, N. H. M., Franceschini-Santos, V. H., Breda, J., Yücel, H., Hernández-Quiles, M., van Lieshout, T., Urzúa-Traslaviña, C. G., Boi, M. C. L., Akbarzadeh, M., et al. (2026) · Nature · doi:10.1038/s41586-025-10093-z · Cited by 33
“As previously described 7,10,23,26 , ISM involves predicting the promoter activity of a reference sequence and sequences with all possible single mutations at the position of interest.”
Maven, B. E. J., Gifford, C. A., Weilert, M., González‐Terán, B., Hüttenhain, R., Pelonero, A., Ivey, K. N., Samse, K. M., Kwong, W., Gordon, D. E., et al. (2023) · Stem Cell Reports · doi:10.1016/j.stemcr.2023.09.014 · Cited by 13
Self-citationHudaiberdiev, S., Taylor, D. L., Song, W., Narisu, N., Bhuiyan, R. M., Taylor, H. J., Tang, X., Yan, T., Swift, A. J., Bonnycastle, L. L., et al. (2023) · Proceedings of the National Academy of Sciences · doi:10.1073/pnas.2206612120 · Cited by 13
“Previous DL methods for TFBS prediction, such as BPNet (43), have used saliency maps to detect TF motifs [e.g., DeepLIFT (44), TF-MoDISco (45)] and applied the detected motifs to predict TF binding.”
Lin, J., Luo, R., & Pinello, L. (2024) · bioRxiv (Cold Spring Harbor Laboratory) · doi:10.1101/2024.08.01.606099 · Cited by 13
Zhou, Y., Adeluwa, T., Zhu, L., Salazar-Magaña, S., Sumner, S., Kim, H., Gona, S., Nyasimi, F. M., Kulkarni, R. N., Powell, J. E., et al. (2025) · Cell Genomics · doi:10.1016/j.xgen.2025.100875 · Cited by 17
DNALONGBENCH: a benchmark suite for long-range DNA prediction tasks
Cheng, W., Song, Z., Zhang, Y., Wang, S., Wang, D., Yang, M., Li, L., & Ma, J. (2025) · Nature Communications · doi:10.1038/s41467-025-65077-4 · Cited by 20
Zhang, Y., Liu, Y., Wang, Z., Wang, M., Xiong, S., Huang, G. C., & Gong, M. (2022) · Genes · doi:10.3390/genes13111952 · Cited by 12
Modeling transcriptional regulation of model species with deep learningPreprint
Cofer, E. M., Raimundo, J., Tadych, A., Yamazaki, Y., Wong, A. K., Theesfeld, C. L., Levine, M., & Troyanskaya, O. G. (2021) · Genome Research · doi:10.1101/gr.266171.120 · Cited by 11
“In particular, the many successive convolutional layers used in these models allow them to learn relatively complex motifs and, we presume, interactions between them (LeCun et al. 2015; Avsec et al. 2021).”
BackgroundMethodDissecting the regulatory logic of specification and differentiation during vertebrate embryogenesisPreprint
Liu, J., Castillo-Hair, S. M., Du, L. Y., Wang, Y., Carte, A. N., Colomer-Rosell, M., Yin, C., Seelig, G., & Schier, A. F. (2024) · bioRxiv (Cold Spring Harbor Laboratory) · doi:10.1101/2024.08.27.609971 · Cited by 14
“Finally, to generate the plots in Figures 3D-F, we realigned motifs to their corresponding cell state-specific peaks using the motif scan method in Avsec et al (Avsec et al., 2021)which considers nucleotide contributions in addition to motif PWM similarity.”
MethodAn intrinsically interpretable neural network architecture for sequence-to-function learning
Balcı, A. T., Ebeid, M. M., Benos, P. V., Kostka, D., & Chikina, M. D. (2023) · Bioinformatics · doi:10.1093/bioinformatics/btad271 · Cited by 15
Unveiling Recent Trends in Biomedical Artificial Intelligence Research: Analysis of Top-Cited Papers
Glicksberg, B. S., & Klang, E. (2024) · Applied Sciences · doi:10.3390/app14020785 · Cited by 12
Gjoni, K., & Pollard, K. S. (2024) · Bioinformatics · doi:10.1093/bioinformatics/btae340 · Cited by 11
“These sequence-to-profile models can predict biological features—including gene expression [Enformer (Avsec et al. 2021a), ExPecto (Zhou et al. 2018), Xpresso (Agarwal and Shendure 2020)], genome folding [Akita (Fudenberg, Kelley and Pollard 2020), C.origami (Tan et al. 2023), DeepC (Schwessinger…”
BackgroundMethodTransfer learning for cross-context prediction of protein expression from 5’UTR sequence
Gilliot, P.-A., & Gorochowski, T. E. (2024) · Nucleic Acids Research · doi:10.1093/nar/gkae491 · Cited by 12
“In relation to this work, approaches focused on feature attribution ( 93 ) and propagation-based methods ( 94 ) that allow for relationships to be inferred between sequence motifs would be relevant for inferring possible mechanistic roles of sub-sequences ( 95 ).”
MethodGauge fixing for sequence-function relationships
(2025) · doi:10.1371/journal.pcbi.1012818 · Cited by 11
Tokenization and deep learning architectures in genomics: A comprehensive review
Testagrose, C., & Boucher, C. A. (2025) · Computational and Structural Biotechnology Journal · doi:10.1016/j.csbj.2025.07.038 · Cited by 13
Tomtom-lite: accelerating Tomtom enables large-scale and real-time motif similarity scoring
Schreiber, J. M. (2025) · Bioinformatics · doi:10.1093/bioinformatics/btaf577 · Cited by 15
“This can be done either by mapping the discrete sequence within seqlets to a motif database, or by mapping a PWM containing these attribution values (sometimes called a contribution-weight matrix or CWM [22]) to a database of attribution-based motifs.”
Wang, Y., Hou, Z., Yang, Y., Wong, K., & Li, X. (2022) · PLoS Computational Biology · doi:10.1371/journal.pcbi.1010779 · Cited by 11
“S2 Table. summarized the comparison of the meaningful motifs detected by SMFM and BPNet, SMFM finally obtained 45 sets of comparison results corresponding to 28 motifs with different IDs in the database, while BPNet obtained 28 sets corresponding to 17 motifs with different IDs.”
MethodInfluentialWu, Y.-H., Huang, Y., Li, J., You, Z., Hu, P., Hu, L., Leung, V. C. M., & Du, Z. (2023) · PLoS Computational Biology · doi:10.1371/journal.pcbi.1011207 · Cited by 11
Chea, S., Kreger, J., Lopez‐Burks, M. E., MacLean, A. L., Lander, A. D., & Calof, A. L. (2024) · Science Advances · doi:10.1126/sciadv.adl4239 · Cited by 13
“We turned to a recent study ( 119 ), in which Avsec et al. captured 52,456 Nanog binding sites in the genome of mESCs, and reported that 3645 genes showed one or more Nanog binding sites located within 250 nucleotides (nt) on either side of their transcription start sites (TSSs) (Fig.”
InfluentialPoised PABP–RNA hubs implement signal-dependent mRNA decay in development
Modic, M., Kuret, K., Steinhauser, S., Faraway, R., van Genderen, E., Mozos, I. R. D. L., Novljan, J., Vičič, Ž., Lee, F. C. Y., ten Berge, D., et al. (2024) · Nature Structural & Molecular Biology · doi:10.1038/s41594-024-01363-x · Cited by 11
“We further investigated which sequence motifs and 3′UTR regions are of high predictive importance by combining Shapley additive explanations (SHAP) 24 and TF-MoDISco (TensorFlow motif discovery and scoring) 25 (see Methods for details).”
InfluentialComputational approaches to understand transcription regulation in development
van der Sande, M., Frölich, S., & van Heeringen, S. J. (2023) · Biochemical Society Transactions · doi:10.1042/bst20210145 · Cited by 12
Characterizing homology-induced data leakage and memorization in genome-trained sequence modelsPreprint
Rafi, A. M., Kiyota, B., Yachie, N., & de Boer, C. G. (2025) · bioRxiv (Cold Spring Harbor Laboratory) · doi:10.1101/2025.01.22.634321 · Cited by 20
Gauge fixing for sequence-function relationships
Posfai, A., Zhou, J., McCandlish, D. M., & Kinney, J. B. (2024) · doi:10.1101/2024.05.12.593772 · Cited by 11
Lee, D. H., Park, E. G., Lee, Y. J., Jeong, H.-S., Roh, H.-Y., Jeong, G.-R., Kim, S.-W., & Kim, H. (2025) · International Journal of Molecular Sciences · doi:10.3390/ijms262210925 · Cited by 13
FUN-PROSE: A deep learning approach to predict condition-specific gene expression in fungi
Nambiar, A., Dubinkina, V. B., Liu, S., & Maslov, S. (2023) · PLoS Computational Biology · doi:10.1371/journal.pcbi.1011563 · Cited by 10
“CNNs were previously applied to extract TF-binding motifs and their higher-order organizational context from ChIP–seq [15, 16], ChIP-exo [17], and artificial sequence experiments [18].”
Murphy, A. E., Askarova, A., Lenhard, B., Skene, N. G., & Marzi, S. J. (2025) · Nucleic Acids Research · doi:10.1093/nar/gkae1212 · Cited by 10
“There has been a recent shift away from arbitrarily benchmarking performance, to prioritizing the use of these models to make new biological discoveries ( 21–23 ).”
EIciRNAs in focus: current understanding and future perspectives
Yang, Y., Zhong, Y., & Chen, L. (2024) · RNA Biology · doi:10.1080/15476286.2024.2443876 · Cited by 10
Interpreting the CTCF-mediated sequence grammar of genome folding with AkitaV2
Smaruj, P. N., Kamulegeya, F., Kelley, D. R., & Fudenberg, G. (2025) · PLoS Computational Biology · doi:10.1371/journal.pcbi.1012824 · Cited by 9
“Recent applications included learning the motif syntax and grammar of transcription factors, such as Nanog [24], OSK, and AP-1 [25].”
BackgroundMethodIn silico prediction of variant effects: promises and limitations for precision plant breeding
Sendrowski, J., Bataillon, T. M., & Ramstein, G. P. (2025) · Theoretical and Applied Genetics · doi:10.1007/s00122-025-04973-1 · Cited by 12
“The state-of-the-art Enformer model is a hybrid CNN which includes trans-former layers to account for nucleotide interactions in long sequence inputs (up to 100 kb, versus 20 kb for Basenji2) (Avsec et al. 2021a).”
InfluentialA scalable approach to investigating sequence-to-function predictions from personal genomes
Spiro, A. E., Tu, X., Sheng, Y., Sasse, A., Hosseini, R., Chikina, M. D., & Mostafavi, S. (2026) · Nature Methods · doi:10.1038/s41592-026-03124-8 · Cited by 15
Lee, D., & Kim, S. (2021) · Korean Journal of Pediatrics · doi:10.3345/cep.2021.01438 · Cited by 10
Multi-layered transcriptional control of cranial neural crest development
Candido-Ferreira, I., Lukoseviciute, M., & Sauka‐Spengler, T. (2022) · Seminars in Cell and Developmental Biology · doi:10.1016/j.semcdb.2022.07.010 · Cited by 8
Selecting deep neural networks that yield consistent attribution-based interpretations for genomics.
Majdandzic, A., Rajesh, C., Tang, A. B., Toneyan, S., Labelson, E., Tripathy, R. K., & Koo, P. K. (2022) · PubMed · Cited by 12
Baumgarten, N., Rumpf, L., Kessler, T., & Schulz, M. H. (2024) · iScience · doi:10.1016/j.isci.2024.109765 · Cited by 10
Kaplow, I. M., Banerjee, A., & Foo, C.-S. (2022) · BMC Genomics · doi:10.1186/s12864-022-08486-9 · Cited by 9
Han, S. K., Muto, Y., Wilson, P. C., Humphreys, B. D., Sampson, M. G., Chakravarti, A., & Lee, D. (2022) · Proceedings of the National Academy of Sciences · doi:10.1073/pnas.2212810119 · Cited by 10
Dissecting regulatory syntax in human development with scalable multiomics and deep learningPreprint
Liu, B. B., Jessa, S., Kim, S. H., Ng, Y. T., Higashino, S. I., Marinov, G. K., Chen, D. C., Parks, B. E., Li, L., Nguyen, T. C., et al. (2025) · bioRxiv (Cold Spring Harbor Laboratory) · doi:10.1101/2025.04.30.651381 · Cited by 10
Self-citationSalvatore, M., Horlacher, M., Marsico, A., Winther, O., & Andersson, R. (2023) · NAR Genomics and Bioinformatics · doi:10.1093/nargab/lqad026 · Cited by 13
“Recent developments to this end (43,18,19,58) show great promise, and we expect that integration of such analyses with the transfer learning scheme of ChromTransfer will be important for future efforts to understand the regulatory code.”
BackgroundThe dynamic landscape of enhancer-derived RNA during mouse early embryo development
Yu, H., Zhao, J., Shen, Y., Qiao, L., Liu, Y., Xie, G., Chang, S., Ge, T., Li, N., Chen, M., et al. (2024) · Cell Reports · doi:10.1016/j.celrep.2024.114077 · Cited by 9
Polygraph: a software framework for the systematic assessment of synthetic regulatory DNA elements
Lal, A., Gunsalus, L. M., Gupta, A., Biancalani, T., & Eraslan, G. (2025) · Genome biology · doi:10.1186/s13059-025-03584-9 · Cited by 12
Zhai, J., Zhang, Y., Zhang, C., Yin, X., Song, M., Tang, C., Ding, P., Li, Z., & Ma, C. (2025) · Advanced Science · doi:10.1002/advs.202503135 · Cited by 9
Mammalian genome writing: Unlocking new length scales for genome engineering
Pinglay, S., Atwater, J. T., Brosh, R., Shendure, J., Maurano, M. T., & Boeke, J. D. (2026) · Cell · doi:10.1016/j.cell.2025.12.034 · Cited by 11
DeepBend: An interpretable model of DNA bendability
Khan, S. R., Sakib, S., Rahman, M. S., & Samee, M. A. H. (2023) · iScience · doi:10.1016/j.isci.2023.105945 · Cited by 9
Optimisation of TP53 reporters by systematic dissection of synthetic TP53 response elements
Trauernicht, M., Rastogi, C., Manzo, S. G., Bussemaker, H., & van Steensel, B. (2023) · bioRxiv · doi:10.1093/nar/gkad718 · Cited by 8
“Periodic helical position effects on transcriptional activity have been observed previously for other TFs ( 22 , 38 , 39 ).”
Quan, L., Mei, J., He, R., Sun, X., Nie, L., Li, K., & Lyu, Q. (2021) · IEEE Journal of Biomedical and Health Informatics · doi:10.1109/jbhi.2021.3058518 · Cited by 9
A Precise Bare Simulation Approach to the Minimization of Some Distances. I. Foundations
Broniatowski, M., & Stummer, W. (2022) · IEEE Transactions on Information Theory · doi:10.1109/tit.2022.3215496 · Cited by 9
Ameen, M., Sundaram, L., Banerjee, A., Shen, M., Kundu, S., Nair, S., Shcherbina, A., Gu, M., Wilson, K. D., Varadarajan, A., et al. (2022) · bioRxiv (Cold Spring Harbor Laboratory) · doi:10.1101/2022.06.29.498132 · Cited by 9
Self-citationAlexandari, A. M., Horton, C. A., Shrikumar, A., Shah, N., Li, E., Weilert, M., Pufall, M. A., Zeitlinger, J., Fordyce, P. M., & Kundaje, A. B. (2023) · bioRxiv (Cold Spring Harbor Laboratory) · doi:10.1101/2023.05.11.540401 · Cited by 11
“To test if this in silico marginalization approach can accurately learn to predict quantitative thermodynamic affinities, we: (1) systematically trained BPNet on ChIP-seq data, (2) applied Affinity Distillation to predict affinities for sequences that had been characterized previously via in vitro thermodynamic assays…”
InfluentialSelf-citationThe regulatory grammar of human promoters uncovered by MPRA-trained deep learningPreprint
Barbadilla-Martínez, L., Klaassen, N. H. M., Franceschini-Santos, V. H., Breda, J., Hernández-Quiles, M., van Lieshout, T., Urzúa-Traslaviña, C. G., Yücel, H., Boi, M. C. L., Hermana-Garcia-Agullo, C., et al. (2024) · bioRxiv (Cold Spring Harbor Laboratory) · doi:10.1101/2024.07.09.602649 · Cited by 9
“PARM consists of a convolutional neural net (CNN), which has previously been used to model similar data structures 15-17 , 19 .”
BackgroundMethodHan, D., Li, Y., Wang, L., Liang, X., Miao, Y., Li, W., Wang, S., & Wang, Z. (2024) · Briefings in Bioinformatics · doi:10.1093/bib/bbae110 · Cited by 10
“Models recently introduced, such as BPNet [26], Leopard [27], FCNA [28], FCNsignal [29] and FCNGRU [30], have demonstrated the capability to achieve base-resolution predictions, which have great potential to assess and interpret variant effects on epigenomic profiles [31].”
BackgroundIdentification, characterization, and design of plant genome sequences using deep learning
Wang, Z., Yuan, H., Yan, J., & Liu, J. (2024) · The Plant Journal · doi:10.1111/tpj.17190 · Cited by 9
Recipes and ingredients for deep learning models of 3D genome folding
Smaruj, P. N., Xiao, Y., & Fudenberg, G. (2025) · Current Opinion in Genetics & Development · doi:10.1016/j.gde.2024.102308 · Cited by 8
gReLU: a comprehensive framework for DNA sequence modeling and design
Lal, A., Gunsalus, L. M., Nair, S., Biancalani, T., & Eraslan, G. (2025) · Nature Methods · doi:10.1038/s41592-025-02868-z · Cited by 22
“Suitable loss functions 7,16,18 are provided for each task (Supplementary Note 1).”
InfluentialMotif-based models accurately predict cell type-specific distal regulatory elements
Cornejo‐Páramo, P., Zhang, X., Louis, L., Yang, Y., Li, Z., Humphreys, D., & Wong, E. S. W. (2025) · Nature Communications · doi:10.1038/s41467-025-65362-2 · Cited by 12
“Applications range from predicting the impact of variant perturbations on TF binding, chromatin state, and chromatin accessibility (e.g., DeepSea 70 , Sei 68 , Basenji 71 , BPNet 15 )tothesequencevocabularyoftranscriptionalinitiationsites 66 , as well as inferring evolutionary conservation of chromatin accessibility 5…”
InfluentialKim, D. S. (2023) · Methods in molecular biology · doi:10.1007/978-1-0716-2899-7_17 · Cited by 8
The evolution and mutational robustness of chromatin accessibility in Drosophila
Khodursky, S., Zheng, E. B., Svetec, N., Durkin, S. M., Benjamin, S., Gadau, A., Wu, X., & Zhao, L. (2023) · Genome biology · doi:10.1186/s13059-023-03079-5 · Cited by 7
“Due to their inspiration from the mammalian visual processing system, convolutional neural networks are able to consider the position and local sequence context while they learn highly complex nonlinear functions.”
Deciphering enhancer sequence using thermodynamics-based models and convolutional neural networks
Dibaeinia, P., & Sinha, S. (2021) · Nucleic Acids Research · doi:10.1093/nar/gkab765 · Cited by 8
“…in contrast to other available implementations of neural networks for regulatory genomics, which are targeted to modeling epigenomic (39,63,64) and genome-wide TF-DNA binding (36,38) data, or do not explicitly model the dependence of sequence function on cellular descriptors such as TF levels (65).”
Rare Does Not Mean Worthless: How Rare Diseases Have Shaped Neurodevelopment Research in the NGS Era
Zaghi, M., Banfi, F., Bellini, E. N., & Sessa, A. (2021) · Biomolecules · doi:10.3390/biom11111713 · Cited by 8
“Thanks to the use of a neural network approach (BPNet, [50]), the authors were able to describe the effect of mutations in a specific cell type using a large single-cell dataset coupled with the Simons Simplex Collection catalog, which comprises over 200,000 pathological variants.”
Method
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
