Pith. sign in

REVIEW 5 major objections 6 minor 38 references

DeepPlantCRE: A Transformer-CNN Hybrid Framework for Plant Gene Expression Modeling and Cross-Species Generalization

T0 review · 5 major / 6 minor · reviewed 2026-08-15 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗ Review JSON

Pith's one-line read DeepPlantCRE predicts plant gene expression from DNA sequence alone and transfers across species with up to 92.3% accuracy.

desk verdict A plausible architecture and a useful multi-species evaluation, but the headline 92.3% cross-species number is not established because the label construction may leak and the text never pins down where the number comes from. read the letter →

arxiv 2505.09883 v1 pith:KACX22GT submitted 2025-05-15 q-bio.GN

Yingjun Wu , Jingyun Huang , Liang Ming , Pengcheng Deng , Maojun Wang , Zeyu Zhang This is my paper · ORCID
classification q-bio.GN
keywords geneexpressionpredictioncis-regulatoryelementsTransformer-CNNhybridcross-speciesgeneralizationplantgenomicstranscriptionfactorbindingsitesDeepLIFTTF-MoDISco
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

This paper proposes DeepPlantCRE, a deep learning model that predicts whether a plant gene is expressed at low, medium, or high level using only the DNA sequence around the gene's start and end. The authors claim the model beats two existing baselines, DeepCRE and PhytoExpr, on accuracy, AUC-ROC, and F1-score across five crop species and eight cotton genomes, and that models trained on one species can predict expression in other species with accuracy up to 92.3%. The aim is to make cis-regulatory element discovery more reliable and transferable, so that regulatory variants and enhancers can be identified from sequence alone and used in crop breeding.

What carries the argument

The central object is the hybrid architecture that places a single Transformer encoder layer (one attention head, embedding dimension 4) before three residual one-dimensional convolutional blocks (kernel size 8; channels 64, 128, 32), followed by a fully connected classification head with sigmoid output. The Transformer first models global sequence context over the entire promoter region; the CNN blocks then extract hierarchical local motif features, and batch normalization, dropout, learning-rate scheduling, and early stopping are used to prevent overfitting. This ordering is the paper's main design claim: global context before local feature extraction yields better cross-species generalization than either architecture alone.

What would settle it

Recompute the same chromosome-held-out cross-validation on one species, but compute the low/medium/high expression cutoffs from the training folds only; if accuracy drops by more than a few points or the ranking of models changes, the reported metrics depend on test-set information. Alternatively, shuffle gene labels and confirm that DeepPlantCRE accuracy approaches chance in the cross-species setting.

Watch

Extended reading notes

Core claim

DeepPlantCRE, a Transformer-CNN hybrid architecture, establishes a new state of the art for sequence-only plant gene expression classification. On five species (Gossypium arboreum, Arabidopsis thaliana, Solanum lycopersicum, Sorghum bicolor, Zea mays) and eight diploid Gossypium genomes, it outperforms DeepCRE and PhytoExpr on accuracy (by 1.6–3.8%), AUC-ROC (by 1.9–3.0%) and F1-score (by 1.2–2.8%), reaching up to 92.3% accuracy in cross-species transfer when Zea mays is the training species. The model's attention mechanism captures long-range regulatory interactions, while residual CNN blocks extract local motifs; interpretability analysis with DeepLIFT and TF-MoDISco recovers known transcription factor binding motifs such as MYR2, TSO1 and TB1, supporting the claim that the model learned biologically meaningful regulatory sequence features.

Load-bearing premise

The evaluation assumes that computing expression cutoffs within each dataset and holding out whole chromosomes for testing yields a fair estimate, with no leakage or distribution shift that would inflate the reported accuracies.

Editorial extensions

If this is right

  • If the model's performance holds, regulatory element discovery in crops can be done directly from promoter sequence, without needing species-specific chromatin data.
  • Cross-species transfer accuracy up to 92.3% suggests a single model can annotate expression-relevant variants across related and even more distant species, prioritizing candidates for gene editing.
  • The recovered motifs matching known TFBSs (MYR2, TSO1, TB1, Zm00001d052229) imply that the model's internal representations align with known plant transcriptional regulation and could nominate novel CREs for validation.
  • The hybrid design's robustness to hyperparameter settings (CNN layers 4–8, kernel size 5–16, learning rate 1e-5 to 1e-3) means the architecture is stable enough for deployment.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • If the chromosome-held-out evaluation is unbiased, the architecture's Transformer-then-CNN ordering could be a reusable pattern for other regulatory genomics tasks where long-range interactions dominate, such as enhancer–promoter interaction prediction.
  • The reported cross-species pattern—maize transfers best, Gossypium worst—suggests training-set size and phylogenetic distance jointly determine transferability; a systematic study across a plant phylogeny could turn this observation into a quantitative transferability predictor.
  • The use of two tissue-specific datasets (leaf) leaves open whether the model generalizes across tissues; a testable extension is to train on multi-tissue expression panels and see whether tissue-specific motifs emerge from the same architecture.
Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Referee Report

5 major / 6 minor

Summary. This manuscript presents DeepPlantCRE, a Transformer-CNN hybrid for classifying plant gene expression from promoter, TSS-proximal, and terminator-proximal DNA sequences. The method is evaluated in within-species chromosome-level cross-validation on five plant species and eight diploid Gossypium genomes, and in cross-species transfer experiments, with reported improvements over DeepCRE and PhytoExpr and a peak cross-species accuracy of 92.3%. The authors also use DeepLIFT and TF-MoDISco to extract sequence motifs and compare them with known transcription factor binding sites in JASPAR.

Significance. The problem of predicting expression from cis-regulatory sequence and testing cross-species generalization is relevant, and the manuscript assembles a broad multi-species benchmark, uses external baselines and an external motif database, and makes code available. These are genuine strengths. However, as presented, the evaluation protocol contains an unresolved label-leakage risk and an ambiguity in the label definition (binary versus three-class), and the cross-species experiment lacks baseline comparisons. The quantitative claims are therefore not yet established, although they are in principle fixable with a revised evaluation.

major comments (5)
  1. [Section 4.1, 4.2] Section 4.1 (Data Processing) says expression labels are assigned by the 25th/75th percentiles of logMaxTPM 'within each dataset,' and Section 4.2 sets k equal to the number of chromosomes. The text never states that the percentile thresholds are recomputed on the training chromosomes only. If the thresholds are estimated from the full species before the split, then the label of every gene on a held-out chromosome is determined using expression values from training chromosomes; this is a preprocessing leak that can inflate Accuracy, AUC-ROC, and F1-score. Please clarify whether thresholds are fold-specific, and if they are not, rerun the evaluation with thresholds computed on training folds only.
  2. [Section 4.3, Figure 3] Section 4.3 reports that the Zea-trained model reaches 92.30% cross-species accuracy, but the text points to no table or figure cell with that value, and Figure 3 is a heatmap with no error bars or numeric annotations. In addition, the target-species labels in the cross-species setting are relative to that species' own logMaxTPM distribution (Section 4.1), so the 92.30% measures consistency of within-species expression ordering, not absolute expression-level prediction. The 'cross-species generalization' interpretation requires either absolute expression labels or an explicit statement that target labels are defined by target-species quantiles; please provide the exact training/test species pair and the precise protocol.
  3. [Section 3.3(iii), 4.1] Section 4.1 creates three expression classes (-1 for low, 0 for medium, 1 for high), while Section 3.3(iii) uses a sigmoid and BCEWithLogitsLoss for binary classification. The treatment of medium-expression genes is never specified: are they excluded, merged with one of the other classes, or modeled as a third class? This ambiguity affects every reported metric and the comparison with DeepCRE and PhytoExpr, so it must be resolved before the results can be interpreted.
  4. [Section 4.3] A stated contribution is improved cross-species generalization (Introduction and RQ2), but Section 4.3 reports transfer results only for DeepPlantCRE; no DeepCRE or PhytoExpr results are shown for the same transfer task. Without matched baselines under the same protocol, the claim that the hybrid architecture generalizes better across species than existing models is not tested.
  5. [Equation (1), Section 4.1] Equation (1) defines the input alphabet as {A,C,G,T,N} but sets X in {0,1}^{L x D} with D equal to the number of standard base types; the representation of N is unspecified. Since plant genome assemblies contain ambiguous bases in promoter and terminator regions, the actual input encoding and any masking or imputation for N must be described for the experiments to be reproducible.
minor comments (6)
  1. [Abstract, Figure 1] The Abstract and Figure 1 list 'Arabidopsis thaliana' twice and omit 'Zea mays' from the species list, even though Zea mays is used in the experiments and is central to the reported 92.3% accuracy; Section 4.1 lists the five species correctly.
  2. [Section 4.1, 4.2] Baseline citations are inconsistent: Section 4.1 identifies DeepCRE as [20] and PhytoExpr as [13], but Section 4.2 refers to 'DeepCRE [13]' and 'PhytoExpr [20]'; please correct the citation mapping.
  3. [Section 4.1] The description of the classification head says the fully connected layer 'eventually output a continuous value representing the predicted gene expression level,' which conflicts with the binary sigmoid classification described in Section 3.3(iii); please harmonize the terminology.
  4. [Figure 2] The Figure 2 caption uses 'unique hot encoding'; this should be 'one-hot encoding'.
  5. [Tables 1 and 2] No statistical significance tests are reported; given the overlapping standard deviations in Tables 1 and 2, the authors should report paired tests or confidence intervals for the main comparisons.
  6. [Section 4.6] The text contains typos such as 'pepole' for 'people' and 'firt' for 'first' in the case study; the manuscript would benefit from a careful proofreading pass.

Circularity Check

1 steps flagged · score 2.0 of 10

No self-derivation circularity; only a label-construction leak in the evaluation protocol keeps this from a clean bill.

  1. other [Section 4.1 (Data Processing) and Section 4.2 (Model Performance), propagated to Section 4.3 (Transfer Learning Study)]
    "For each species, gene expression levels were categorized into low, medium and high groups based on the distribution of the logMaxTPM values < 25% marked as -1 (low expression), 25%-75% as 0 (medium expression), and > 75% as 1 (high expression). ... For each species, we employed k fold cross-validation for training, with k setting as the number of chromosomes of each species under training."

    The quantile thresholds are computed once per species before the chromosome-level splits. Under k-fold CV with k equal to the number of chromosomes, held-out chromosome labels are therefore derived from expression values that include training chromosomes. Reported accuracy, AUC-ROC, and F1-score on held-out chromosomes partly reflect how well the training-set quantiles reproduce the held-out ternary labels, rather than purely how well DNA sequence predicts expression class. The same issue extends to the cross-species experiment, where target-species labels are relative to each target species' own full distribution.

full rationale

DeepPlantCRE's derivation chain is otherwise self-contained: the model is defined by standard one-hot encoding, Transformer and CNN residual blocks, BCE loss, and Adam optimization (Eqs. 1-13), with no load-bearing self-citation or imported uniqueness theorem. The performance claims are benchmarked against external baselines (DeepCRE, PhytoExpr), and the interpretability claim is validated against the external JASPAR database and a known maize TFBS with reported p/e values. No equation reduces to its own inputs, and no fitted parameter is renamed as a prediction. The only circularity-adjacent concern is the label construction: percentile thresholds are computed per species before the chromosome-level train/test split, so held-out labels are not independent of training data. This can inflate absolute metrics, especially the headline 92.3% cross-species accuracy, but it does not make the model's output equivalent to its input, and the comparison to baselines under the same protocol retains relative value. Score 2 reflects this minor evaluation leak rather than a fundamental circular derivation.

Assumptions & free parameters 4 free parameters · 3 assumptions · 0 invented entities

The model does not introduce new biological entities. Its free parameters are the usual architectural choices plus the expression-percentile cutoffs that define the classification task. The key domain assumption is that a promoter-proximal window is enough to predict expression level, a simplification that is also true of the baseline models being compared.

free parameters (4)
  • Expression cutoffs (25th and 75th percentiles of logMaxTPM) = thresholds at the 25th and 75th percentiles of each species' expression distribution
    The ternary class labels (low, medium, high) are defined by these cutoffs, so the classification target is a discretization of the same data used to evaluate the model.
  • Transformer embedding dimension d = stated as d = 4 in Section 3.5 and as matching the one-hot alphabet in Section 4.1
    The choice of a 4-dimensional embedding for a 4-letter alphabet makes the attention projections very low-rank; this is a hand-chosen architectural constant that the paper does not justify.
  • CNN channel schedule = 64, 128, 32
    Channel counts in the three residual blocks are chosen by hand; the paper reports robustness to layer count but does not test channel widths.
  • Hyperparameters (kernel size 8, dropout 0.25, learning rate 1e-4, early stopping patience 10) = kernel=8, p=0.25, lr=1e-4, patience=10
    These are standard choices validated by sensitivity experiments on a single species, but they remain tuning choices that the final results depend on.
assumptions (3)
  • domain assumption DNA sequence windows around TSS and TTS are sufficient to predict expression class.
    The model uses only 1 kb upstream to 0.5 kb downstream of TSS and 0.5 kb upstream to 1 kb downstream of TTS; long-range enhancers beyond these windows cannot influence predictions, which is the same limitation the paper attributes to CNN-only models.
  • domain assumption The hierarchical ordering (Transformer followed by CNN) is a valid way to combine global and local sequence features.
    The paper motivates this ordering but provides no architectural search or comparison with CNN-first or parallel hybrids.
  • domain assumption DeepLIFT importance scores, TF-MoDISco motifs, and MEME similarity to JASPAR correctly identify the biological features the model uses.
    The interpretability pipeline is standard, but the paper does not validate how faithfully the extracted motifs explain the model's predictions, nor how many motifs matched known TFBSs out of the total discovered.

how reviews work

0 comments
Cite this review

Pith. "Pith review of DeepPlantCRE: A Transformer-CNN Hybrid Framework for Plant Gene Expression Modeling and Cross-Species Generalization." pith.science (2026). https://pith.science/paper/KACX22GT

@misc{pith2026250509883,
  author       = {Pith},
  title        = {Pith review of: DeepPlantCRE: A Transformer-CNN Hybrid Framework for Plant Gene Expression Modeling and Cross-Species Generalization},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/KACX22GT}},
  note         = {Machine review of arXiv:2505.09883}
}
read the original abstract

The investigation of plant transcriptional regulation constitutes a fundamental basis for crop breeding, where cis-regulatory elements (CREs), as the key factor determining gene expression, have become the focus of crop genetic improvement research. Deep learning techniques, leveraging their exceptional capacity for high-dimensional feature extraction and nonlinear regulatory relationship modeling, have been extensively employed in this field. However, current methodologies present notable limitations: single CNN-based architectures struggle to capture long-range regulatory interactions, while existing CNN-Transformer hybrid models demonstrate proneness to overfitting and inadequate generalization in cross-species prediction contexts. To address these challenges, this study proposes DeepPlantCRE, a deep-learning framework for plant gene expression prediction and CRE Extraction. The model employs a Transformer-CNN hybrid architecture that achieves enhanced Accuracy, AUC-ROC, and F1-score metrics over existing baselines (DeepCRE and PhytoExpr), with improved generalization performance and overfitting inhibiting. Cross-species validation experiments conducted on gene expression datasets from \textit{Gossypium}, \textit{Arabidopsis thaliana}, \textit{Solanum lycopersicum}, \textit{Sorghum bicolor}, and \textit{Arabidopsis thaliana} reveal that the model achieves peak prediction accuracy of 92.3\%, particularly excelling in complex genomic data analysis. Furthermore, interpretability investigations using DeepLIFT and Transcription Factor Motif Discovery from the importance scores algorithm (TF-MoDISco) demonstrate that the derived motifs from our model exhibit high concordance with known transcription factor binding sites (TFBSs) such as MYR2, TSO1 in JASPAR plant database, substantiating the potential of biological interpretability and practical agricultural application of DeepPlantCRE.

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

38 extracted references · 36 canonical work pages

  1. [1]

    Adamo, M

    M. Adamo, M. Chialva, J. Calevo, F. Bertoni, K. Dixon, and S. Mam- mola. Plant scientists’ research attention is skewed towards colourful, conspicuous and broadly distributed flowers. Nature plants, 7(5):574– 578, 2021

    work page 2021

  2. [2]

    Akagi, K

    T. Akagi, K. Masuda, E. Kuwada, K. Takeshita, T. Kawakatsu, T. Ari- izumi, Y . Kubo, K. Ushijima, and S. Uchida. Genome-wide cis- decoding for expression design in tomato using cistrome data and ex- plainable deep learning. The Plant Cell, 34(6):2174–2187, 2022

    work page 2022

  3. [3]

    Alipanahi, A

    B. Alipanahi, A. Delong, M. T. Weirauch, and B. J. Frey. Predicting the sequence specificities of dna-and rna-binding proteins by deep learning. Nature biotechnology, 33(8):831–838, 2015

    work page 2015

  4. [4]

    V . Ashish. Attention is all you need. Advances in neural information processing systems, 30:I, 2017

    work page 2017

  5. [5]

    Avsec, V

    Ž. Avsec, V . Agarwal, D. Visentin, J. R. Ledsam, A. Grabska- Barwinska, K. R. Taylor, Y . Assael, J. Jumper, P. Kohli, and D. R. Kel- ley. Effective gene expression prediction from sequence by integrating long-range interactions. Nature methods, 18(10):1196–1203, 2021

    work page 2021

  6. [6]

    A. M. Bolger, M. Lohse, and B. Usadel. Trimmomatic: a flexible trim- mer for illumina sequence data. Bioinformatics, 30(15):2114–2120, 2014

    work page 2014

  7. [7]

    J. A. Castro-Mondragon, R. Riudavets-Puig, I. Rauluseviciute, R. Berhanu Lemma, L. Turchi, R. Blanc-Mathieu, J. Lucas, P. Boddie, A. Khan, N. Manosalva Pérez, et al. Jaspar 2022: the 9th release of the open-access database of transcription factor binding profiles. Nucleic acids research, 50(D1):D165–D173, 2022

    work page 2022

  8. [8]

    S. K. Govers, M. Campos, B. Tyagi, G. Laloux, and C. Jacobs-Wagner. Apparent simplicity and emergent robustness in the control of the es- cherichia coli cell cycle. Cell systems, 15(1):19–36, 2024

    work page 2024

Show all 38 references
  1. [9]

    Gupta, J

    S. Gupta, J. A. Stamatoyannopoulos, T. L. Bailey, and W. S. Noble. Quantifying similarity between motifs. Genome biology, 8:1–9, 2007

    2007

  2. [10]

    H. Jung, R. Covino, A. Arjun, P. G. Bolhuis, and G. Hum- mer. Autonomous artificial intelligence discovers mechanisms of molecular self-organization in virtual experiments. arXiv preprint arXiv:2105.06673, 2021

    2021 arXiv

  3. [11]

    D. Kim, B. Langmead, and S. L. Salzberg. Hisat: a fast spliced aligner with low memory requirements. Nature methods, 12(4):357–360, 2015

    2015

  4. [12]

    H. Li, B. Handsaker, A. Wysoker, T. Fennell, J. Ruan, N. Homer, G. Marth, G. Abecasis, R. Durbin, and . G. P. D. P. Subgroup. The sequence alignment/map format and samtools. bioinformatics, 25(16): 2078–2079, 2009

    2009

  5. [13]

    T. Li, H. Xu, S. Teng, M. Suo, R. Bahitwa, M. Xu, Y . Qian, G. P. Ram- stein, B. Song, E. S. Buckler, et al. Modeling 0.6 million genes for the rational design of functional cis-regulatory variants and de novo design of cis-regulatory sequences. Proceedings of the National Aca...

    2024

  6. [14]

    X. Li, T. Liang, and H. Liu. How plants coordinate their development in response to light and temperature signals. The Plant Cell , 34(3): 955–966, 2022

    2022

  7. [15]

    Maghraby and M

    A. Maghraby and M. Alzalaty. Genome-wide identification and evolu- tionary analysis of the ap2/erebp, cox and ltp genes in zea mays l. under drought stress. Scientific Reports, 14(1):7610, 2024

    2024

  8. [16]

    Muller, J

    H. Muller, J. Gil, and I. A. Drinnenberg. The impact of centromeres on spatial genome architecture. Trends in Genetics, 35(8):565–578, 2019

    2019

  9. [17]

    Nguyen, D

    V .-A. Nguyen, D. W. Bartels, and C. A. Gilligan. Modelling the spread and mitigation of an emerging vector-borne pathogen: Citrus greening in the us. PLOS Computational Biology, 19(6):e1010156, 2023

    2023

  10. [18]

    J. Ou, S. A. Wolfe, M. H. Brodsky, and L. J. Zhu. motifstack for the analysis of transcription factor binding site evolution. Nature methods, 15(1):8–9, 2018

    2018

  11. [19]

    Ouyang, F

    Z. Ouyang, F. Liu, W. Li, J. Wang, B. Chen, Y . Zheng, Y . Li, H. Tao, X. Xu, C. Li, et al. The developmental and evolutionary characteris- tics of transcription factor binding site clustered regions based on an explainable machine learning model. Nucleic Acids Research, 52(13...

    2024

  12. [20]

    F. F. Peleke, S. M. Zumkeller, M. Gültas, A. Schmitt, and J. Szyma´nski. Deep learning the cis-regulatory code for gene expression in selected model plants. Nature Communications, 15(1):3488, 2024

    2024

  13. [21]

    Pertea, G

    M. Pertea, G. M. Pertea, C. M. Antonescu, T.-C. Chang, J. T. Mendell, and S. L. Salzberg. Stringtie enables improved reconstruction of a tran- scriptome from rna-seq reads. Nature biotechnology, 33(3):290–295, 2015

    2015

  14. [22]

    Pinto-Cámara, A

    R. Pinto-Cámara, A. Linares, D. Moreno-Gutiérrez, H. O. Hernández, J. Martínez-Reyes, J. Rendón-Mancha, C. D. Wood, and A. Guerrero. Fcslib: an open-source tool for fluorescence fluctuation spectroscopy analysis for mobility, number and molecular brightness in r. Bioinfor- mat...

    1930

  15. [23]

    Rodríguez-Leal, Z

    D. Rodríguez-Leal, Z. H. Lemmon, J. Man, M. E. Bartlett, and Z. B. Lippman. Engineering quantitative trait variation for crop improvement by genome editing. Cell, 171(2):470–480, 2017

    2017

  16. [24]

    Rychel, A

    K. Rychel, A. V . Sastry, and B. O. Palsson. Machine learning uncovers independently regulated modules in the bacillus subtilis transcriptome. Nature communications, 11(1):6338, 2020

    2020

  17. [25]

    Shrikumar, P

    A. Shrikumar, P. Greenside, and A. Kundaje. Learning important fea- tures through propagating activation differences. In International con- ference on machine learning, pages 3145–3153. PMlR, 2017

    2017

  18. [26]

    Shrikumar, K

    A. Shrikumar, K. Tian, Ž. Avsec, A. Shcherbina, A. Banerjee, M. Sharmin, S. Nair, and A. Kundaje. Technical note on transcription factor motif discovery from importance scores (tf-modisco) version 0.5. 6.5. arXiv preprint arXiv:1811.00416, 2018

    2018 arXiv

  19. [27]

    J.-Y . Song, T. Leung, L. K. Ehler, C. Wang, and Z. Liu. Regulation of meristem organization and cell division by tso1, an arabidopsis gene with cysteine-rich repeats. Development, 127(10):2207–2217, 2000

    2000

  20. [28]

    Takeda, Y

    T. Takeda, Y . Suwa, M. Suzuki, H. Kitano, M. Ueguchi-Tanaka, M. Ashikari, M. Matsuoka, and C. Ueguchi. The ostb1 gene negatively regulates lateral branching in rice. The Plant Journal, 33(3):513–520, 2003

    2003

  21. [29]

    X. Tu, M. K. Mejía-Guerra, J. A. Valdes Franco, D. Tzeng, P.-Y . Chu, W. Shen, Y . Wei, X. Dai, P. Li, E. S. Buckler, et al. Reconstructing the maize leaf regulatory network using chip-seq data of 104 transcription factors. Nature communications, 11(1):5089, 2020

    2020

  22. [30]

    Vaswani, N

    A. Vaswani, N. Shazeer, N. Parmar, J. Uszkoreit, L. Jones, A. N. Gomez, Ł. Kaiser, and I. Polosukhin. Attention is all you need. Ad- vances in neural information processing systems, 30, 2017

    2017

  23. [31]

    L. Wang, Y . Xi, S. Sung, and H. Qiao. Rna-seq assistant: machine learning based methods to identify more transcriptional regulated genes. BMC genomics, 19:1–13, 2018

    2018

  24. [32]

    M. Wang, J. Li, Z. Qi, Y . Long, L. Pei, X. Huang, C. E. Grover, X. Du, C. Xia, P. Wang, et al. Genomic innovation and regulatory rewiring during evolution of the cotton genus gossypium. Nature Genetics, 54 (12):1959–1971, 2022

    1959

  25. [33]

    Z. Wang, Y . Peng, J. Li, J. Li, H. Yuan, S. Yang, X. Ding, A. Xie, J. Zhang, S. Wang, et al. Deepcba: A deep learning framework for gene expression prediction in maize based on dna sequences and chromatin interactions. Plant Communications, 5(9), 2024

    2024

  26. [34]

    L. Wu, Z. Luo, Y . Shi, Y . Jiang, R. Li, X. Miao, F. Yang, Q. Li, H. Zhao, J. Xue, et al. A cost-effective tscut&tag method for profiling transcrip- tion factor binding landscape. Journal of Integrative Plant Biology, 64 (11):2033–2038, 2022

    2022

  27. [35]

    W. Yan, Z. Li, C. Pian, and Y . Wu. Plantbind: an attention-based multi- label neural network for predicting plant transcription factor binding sites. Briefings in Bioinformatics, 23(6):bbac425, 2022

    2022

  28. [36]

    J. You, Z. Liu, Z. Qi, Y . Ma, M. Sun, L. Su, H. Niu, Y . Peng, X. Luo, M. Zhu, et al. Regulatory controls of duplicated gene expression dur- ing fiber development in allotetraploid cotton. Nature genetics, 55(11): 1987–1997, 2023

    1987

  29. [37]

    C. Zhao, A. Hanada, S. Yamaguchi, Y . Kamiya, and E. P. Beers. The arabidopsis myb genes myr1 and myr2 are redundant negative regula- tors of flowering time under decreased light intensity. The Plant Jour- nal, 66(3):502–515, 2011

    2011

  30. [38]

    P. Zhou, T. A. Enders, Z. A. Myers, E. Magnusson, P. A. Crisp, J. M. Noshay, F. Gomez-Cano, Z. Liang, E. Grotewold, K. Greenham, et al. Prediction of conserved and variable heat and cold stress response in maize using cis-regulatory information. The Plant Cell, 34(1):514–534, 2022

    2022

Pith tools

Reviewed August 15, 2026 · model on record in the stance chip above.