Pith. sign in

REVIEW 3 major objections 5 minor 50 references

MSGR argues that spatial gene expression prediction improves when target genes are decoded coarse-to-fine along a Gene Ontology tree rather than as a flat vector, and that the gain comes from the biological structure itself, not from hierar

Reviewed by Pith at T0; open to challenge. T0 means a machine referee read the full paper against a public rubric. the ladder, T0–T4 →

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

MSGR's Gene Ontology-guided hierarchical decoder improves spatial gene expression prediction from histology images, with the biological structure adding a +0.027 gain over an equivalent random hierarchy.

T0 review reviewed 2026-08-05 challenge →

load-bearing objection Solid empirical paper with a real new idea and a good random-hierarchy control; the GO-structure claim holds, but the plug-in compatibility evidence is confounded and needs a matched-capacity control. the 3 major comments →

arxiv 2608.00405 v1 pith:KKIKSGKZ submitted 2026-08-01 cs.AI q-bio.GN

Gene Ontology-Guided Hierarchical Spatial Gene Expression Prediction from Histopathology Images

Zhiwen Xu , Xiaoming Yan , Chengkun Wu , Juan Chen , Haoang Chi , Liyang Xu This is my paper
classification cs.AI q-bio.GN
keywords spatial transcriptomicshistopathology imagesGene Ontologyhierarchical decodingresidual refinementgene expression predictionplug-in decoderHEST-1k
verification ladder T0 review T1 audit T2 compute T3 formal T4 reserved

The pith

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

The reading

The paper argues that predicting spatial gene expression from histology images should not treat the target genes as an unstructured 200-dimensional output. It organizes the genes into a four-level tree derived from Gene Ontology (virtual root, GO domains, GO terms, individual genes) and decodes coarse-to-fine, with each level adding a residual correction to the inherited parent prediction under scale-weighted supervision. On nine HEST-1k datasets, the GO-guided decoder reports an average PCC-200 of 0.517, ahead of flat decoders including the generative baseline STFlow (0.503). A structurally identical random hierarchy scores 0.490, a 0.027 gap that the paper attributes to the biological content of the ontology rather than to hierarchy depth or shape. Because the decoder only replaces the output head, it can be inserted into existing pipelines and the paper reports consistent improvements for STFlow, EGN, and ST-Net (+0.019, +0.017, +0.018 average PCC-200).

Core claim

MSGR's central claim is that a panel of hundreds of genes can be predicted from an H&E image by first forming a coarse functional picture and then refining it. The target genes are arranged offline into a four-level GO tree; the expression target of every non-leaf node is the mean expression of its descendant genes. A decoder shared across levels starts from the virtual root, and at each finer level the parent prediction is broadcast to children, a residual correction is learned from conditioning tokens, and latent context from the parent scale is propagated forward through a cross-attention highway. The paper's evidence that this structure matters: the same backbone without the hierarchy (S

What carries the argument

The central object is the four-level GO tree projection: virtual root, up to four level-1 nodes (three GO domains plus an unannotated proxy), a level-2 layer of GO terms or proxies, and level-3 leaf genes, with each gene assigned exactly once to its most specific annotation and each GO term connected to its largest parent. Coarse targets are means over descendant leaves, so supervision is consistent across scales. The mechanism that carries the argument is the shared multi-scale residual decoder: scale-0 outputs a root estimate; every finer scale broadcasts the parent prediction to its children and adds a residual from a shared AdaLN Transformer block, while a latent highway concatenates par

Load-bearing premise

The projection of the GO DAG into a single-parent tree must retain enough functional coherence that the coarse-scale supervision targets are biologically meaningful; if that projection scrambles the signal, the method would reduce to an arbitrary hierarchy.

What would settle it

Train the identical MSGR decoder on a hierarchy in which GO term assignments have been randomly permuted across genes while preserving tree shape and term sizes. If the PCC-200 gap over flat decoding does not collapse toward zero, the claimed biological-structure contribution is not doing the work.

Watch this falsifier. Get emailed when new claim-graph text bears on it.

If this is right

  • The GO-guided decoder can replace the final linear head of diverse models, with average PCC-200 gains of +0.019 for STFlow, +0.017 for EGN, and +0.018 for ST-Net across nine datasets.
  • GO-structured decoding improves the top-10, top-50, and top-200 gene correlations, with reported averages of 0.776, 0.710, and 0.517, respectively.
  • Intermediate GO-term predictions constitute interpretable outputs that align with independent pathway activity annotations, with a reported mean correlation of 0.896 on SKCM.
  • The gene-side decoder is orthogonal to image-side multi-magnification methods, so both enhancement strategies can be combined in the same architecture.
  • Both cross-scale communication (latent highway, -0.015) and residual inheritance (-0.013) contribute to the final average PCC-200, so the benefit is not from the tree alone but from the coarse-to-fine refinement protocol.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

  • If the biological-structure effect is genuine, then other functional gene groupings—pathway modules, protein interaction communities, even data-driven co-expression modules—should also help when used as the hierarchy; the paper's own alternative-prior ablation shows they help less but still beat a random tree, suggesting the benefit comes from coherent functional grouping rather than from GO's cur
  • A natural next test is a full DAG-aware decoder that retains gene-to-term multi-parent edges and term-to-term connectivity; the paper's multi-parent averaging variant is not that test and loses 0.017, so it does not settle whether richer topology would help or hurt.
  • Since the plug-in gains are similar across a CNN, a Transformer regressor, and a flow-matching model, the same gene-side hierarchy should transfer to other multi-gene prediction heads, such as super-resolution expression or cell-type deconvolution, without retraining the image-side modules.
  • A complementary experiment would hold the decoder fixed and build the hierarchy from expression data per tissue instead of from GO; if a learned hierarchy matches or exceeds GO, it would measure how much curated biological prior remains useful once data are abundant.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

3 major / 5 minor

Summary. The paper proposes MSGR, a hierarchical decoder for predicting spatial gene expression from histopathology images. It projects the Gene Ontology DAG into a four-level tree (root, domains, terms, genes) and uses a shared Transformer backbone to produce residual corrections from coarse to fine scales, supervised by multi-scale MSE targets derived from descendant-mean aggregation. The authors claim two central results: (1) GO-structured decoding outperforms flat decoding, including a state-of-the-art generative baseline (STFlow), across nine HEST-1k datasets; and (2) the gain comes from biological ontology structure rather than hierarchy per se, based on a +0.027 average PCC-200 margin over MSGR-Random, a structurally equivalent random hierarchy. They also claim the decoder is a model-agnostic plug-in that improves STFlow, EGN, and ST-Net by average gains of +0.019, +0.017, and +0.018. The paper includes ablations for the latent highway, residual inheritance, loss weights, alternative biological priors, and single-parent versus multi-parent projections.

Significance. If the claims hold, the paper makes a useful contribution by introducing a biologically structured decoding prior into a task where existing methods treat the gene output space as flat. The random-hierarchy control and the comparison against alternative biological priors (pathway, PPI, coexpression) are well-designed and provide credible evidence that the structured decoding interface itself is not the sole source of the gain for the full MSGR model. The authors also ship source code, which supports reproducibility. However, the plug-in evidence, a central advertised strength, is currently confounded by architectural and objective mismatches, and the main empirical comparisons lack statistical significance testing despite the small reported margins.

major comments (3)
  1. [Section 4.3 vs. Appendix E (Table 2)] The plug-in claim is not cleanly attributable to GO structure. Section 4.3 says modifications are 'strictly limited' to replacing the linear output head and substituting the loss, but Appendix E states that the plug-in uses 'a lightweight variant of the GO-guided decoder based on MLP layers rather than the full Transformer backbone used in MSGR.' For STFlow, the base training objective is flow matching, not 'standard MSE loss,' so swapping to multi-scale MSE changes the training objective. Each plug-in comparison therefore varies head architecture, capacity, and (for STFlow) the learning objective simultaneously. There is no flat-head control with matched capacity and the same loss. The observed average gains of +0.019/+0.017/+0.018, with per-dataset negatives (e.g., LUNG -0.004 for STFlow+MS, PRAD -0.035 for EGN+MS), are thus consistent with a capacity or objective effect. Please report
  2. [Tables 1–3] The central empirical claims are supported only by point estimates. Most reported differences are small (average PCC-200 gains of 0.014–0.027, with several per-dataset deltas near zero or negative), and the number of cross-validation folds is only 2–6 per dataset. The paper reports mean±std across folds but provides no paired significance tests, confidence intervals, or effect-size/rules-of-three analysis. The phrase 'consistently outperforms' is stronger than what the evidence supports. Please add per-dataset significance tests (e.g., paired bootstrap or Wilcoxon across folds) or otherwise quantify uncertainty in the average margins.
  3. [Section 4.5, Table 3 vs. Section 4.3, Table 2] The random-hierarchy control is convincing for the standalone MSGR model but does not validate the lightweight plug-in variant. Table 3 shows MSGR-Random trailing MSGR by 0.027 under the full Transformer backbone, but Table 2's plug-in variants use the Appendix E MLP-based decoder, for which no random-hierarchy ablation is reported. The claim that the plug-in gains are 'attributable to biological ontology structure rather than hierarchical decomposition per se' therefore lacks a control for the plug-in setting. Either add an MSGR-Random version of the lightweight decoder for the three base models, or state the plug-in results as compatibility evidence only, without the biological-attribution claim.
minor comments (5)
  1. [Figure 5 caption] Typo: 'Spaital Transformer' should be 'Spatial Transformer.'
  2. [Table 5] The organ column lists 'KIDNEY' for both CCRCC and KIDNEY rows; please distinguish renal cell carcinoma versus normal kidney or clarify the dataset naming.
  3. [Appendix G.1] Equation (20) is stated but not numbered; please number it for cross-referencing in the discussion of residual inheritance.
  4. [Section 4.2] The claim that MSGR 'outperforms STFlow on eight of nine datasets' should be accompanied by the fold-level standard deviations for those specific comparisons; Table 1 already includes them, but a direct paired-error visualization would help.
  5. [Appendix C.1] The loss-weight ablation is only reported on five datasets; specify whether the remaining four datasets were omitted for computational reasons and note the limitation in the text.

Circularity Check

0 steps flagged

No significant circularity: GO prior is external, random-hierarchy control is a proper counterfactual, and no equation reduces a prediction to a fitted input.

full rationale

The central derivation is self-contained and empirically controlled. The GO hierarchy is constructed offline from g:Profiler GO annotations (Section 3.4), independently of the expression values being predicted; the coarse-scale supervision targets y^(k)_j = mean descendant expression (Eq. 3) are standard hierarchical targets, not fitted parameters renamed as predictions. The MSGR-Random ablation (Section 4.5) replaces the GO tree with a randomly constructed hierarchy of identical shape while holding all other components constant, directly controlling for hierarchical decomposition per se; the +0.027 margin is therefore evidence for biological structure rather than a tautology. Appendix C.2 extends this with pathway, PPI, coexpression, and random priors under the same backbone, further supporting the attribution. No load-bearing self-citation chain or imported uniqueness theorem is present; the DAG-to-tree projection cites external tools, and the method is benchmarked against external baselines on HEST-1k. The plug-in experiments (Section 4.3, Appendix E) change more than the GO structure—the lightweight MLP decoder adds capacity and for STFlow the loss changes from flow matching to MSE—so the plug-in gain is confounded, but this is a validity/correctness concern, not a circularity: no fitted value is claimed as an independent prediction. The Appendix F.1 GO-term/Hallmark alignment is a soft self-consistency check because GO-term predictions are supervised with means of descendant genes that overlap the Hallmark gene sets, weakening the word 'independent,' but it is not load-bearing for the main predictive claims. Overall, the core derivation does not reduce to its inputs by construction.

Axiom & Free-Parameter Ledger

5 free parameters · 7 axioms · 0 invented entities

The central claim rests on the choice of GO as the prior, the tree projection, and hyperparameters selected on the evaluation datasets. No new physical entities are introduced; the virtual root and proxy nodes are bookkeeping devices in the tree construction.

free parameters (5)
  • Multi-scale loss weights = [0,1,2,4]
    Selected in Appendix C.1 by comparing four configurations on five datasets; the adopted weights remove root supervision and progressively emphasize finer scales. This is a hand-tuned hyperparameter, not derived from first principles.
  • Spatial neighborhood size n = 8
    Chosen from the 4x4 sensitivity grid in Section 4.6 and Appendix D; n=8 with L=2 forms the best plateau. Affects the condition processor.
  • Shared Transformer depth L = 2
    Chosen from the sensitivity grid; deeper layers give no consistent gain.
  • Embedding dimension d = 512
    Set in Section 4.1.4 without ablation; a capacity hyperparameter.
  • GO hierarchy construction thresholds = most-specific term; largest parent term; path compression
    Algorithmic choices in Section 3.4 that determine tree shape; not swept, but they define the structural prior.
axioms (7)
  • domain assumption H&E histology images contain sufficient morphological signal to predict spatial gene expression.
    The entire task premise, stated in Section 1 and implicit in all baselines.
  • domain assumption Genes sharing granular GO annotations tend to exhibit correlated expression patterns.
    Section 1, citing [35]; this justifies using GO as a structural prior.
  • domain assumption The GO DAG can be projected to a single-parent four-level tree without losing the functional signal needed for coarse-to-fine prediction.
    Section 3.4; the projection chooses most-specific terms and broadest parents. Appendix F.2 tests a multi-parent variant and finds single-parent better, but the full DAG is not evaluated.
  • domain assumption Mean expression of descendant leaf genes is a meaningful coarse supervision target.
    Equation (3) defines coarser targets as means; Appendix G.2 acknowledges this can attenuate sparse pathway signals.
  • domain assumption The frozen UNI encoder provides transferable visual features for this task.
    Section 3.2 and baselines; UNI pretrained on WSIs.
  • standard math E(2)-equivariant frame averaging yields rotation/reflection-invariant spatial context.
    Appendix B, Equations (13)-(17); this is standard frame averaging [28].
  • domain assumption Transformer self-attention and AdaLN layers can learn the required residual corrections.
    Model architecture in Section 3.5; not proven, empirically validated.

reviewed 2026-08-05 · how reviews work

0 comments
Cite this review

Pith. "Pith review of Gene Ontology-Guided Hierarchical Spatial Gene Expression Prediction from Histopathology Images." pith.science (2026). https://pith.science/paper/KKIKSGKZ

@misc{pith2026260800405,
  author       = {Pith},
  title        = {Pith review of: Gene Ontology-Guided Hierarchical Spatial Gene Expression Prediction from Histopathology Images},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/KKIKSGKZ}},
  note         = {Machine review of arXiv:2608.00405}
}
Share X Bluesky LinkedIn Reddit HN
read the original abstract

Predicting spatial gene expression from histopathology images enables large-scale transcriptomic profiling without the cost of direct measurement. Existing methods decode the target gene set as a flat, unstructured vector, ignoring the inter-gene dependencies arising from shared biological pathways and regulatory programs. Without explicit structural guidance, models must infer these dependencies entirely from limited paired data, constraining prediction quality. We propose MSGR (Multi-Scale Gene Refiner), which bridges this gap by incorporating the Gene Ontology (GO), a curated functional hierarchy of genes, as an explicit structural prior. MSGR organizes target genes into a four-level GO tree. Its GO-guided decoder then progressively refines predictions from coarse functional domains to fine individual genes via residual corrections under scale-weighted supervision. Operating solely on the gene side, the GO-guided decoder serves as a seamless plug-in replacement that consistently improves existing architectures without requiring any image-side modifications. Extensive experiments on nine datasets from the HEST-1k benchmark provide empirical evidence for two central claims: GO-structured decoding consistently outperforms flat decoding, even against a state-of-the-art generative baseline, and the gain is attributable to biological ontology structure rather than hierarchical decomposition per se, as confirmed by a +0.027 margin over a structurally equivalent random hierarchy.

Figures

Figures reproduced from arXiv: 2608.00405 by Chengkun Wu, Haoang Chi, Juan Chen, Liyang Xu, Xiaoming Yan, Zhiwen Xu.

Figure 1
Figure 1. Figure 1: Traditional flat decoding vs. hierarchical decoding [PITH_FULL_IMAGE:figures/full_fig_p002_1.png] view at source ↗
Figure 2
Figure 2. Figure 2: Overview of the MSGR framework. (a) Condition Acquisition & Processing: A frozen UNI encoder extracts per-spot [PITH_FULL_IMAGE:figures/full_fig_p004_2.png] view at source ↗
Figure 3
Figure 3. Figure 3: Predicted spatial expression of MLANA on slide TENX115 from the SKCM dataset, compared against the ground truth. Each panel title reports the per-slide PCC. The color scale encodes the normalized expression level (blue: low, red: high). the GO-guided decoder is compatible with generative, exemplar￾guided, and CNN-based architectures. Integration details and the computational overhead of the lightweight plu… view at source ↗
Figure 4
Figure 4. Figure 4: Hyperparameter sensitivity analysis of the pro [PITH_FULL_IMAGE:figures/full_fig_p008_4.png] view at source ↗
Figure 5
Figure 5. Figure 5: Spatial Transformer architecture. (a) Overall archi [PITH_FULL_IMAGE:figures/full_fig_p010_5.png] view at source ↗
Figure 6
Figure 6. Figure 6: Per-dataset hyperparameter sensitivity across five evaluation datasets. Full [PITH_FULL_IMAGE:figures/full_fig_p015_6.png] view at source ↗
Figure 7
Figure 7. Figure 7: GO hierarchy constructed for the IDC dataset. Node size is proportional to the number of annotated genes. Four [PITH_FULL_IMAGE:figures/full_fig_p016_7.png] view at source ↗

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Reference graph

Works this paper leans on

50 extracted references · 32 canonical work pages

  1. [1]

    Michael Ashburner, Catherine A Ball, Judith A Blake, David Botstein, Heather Butler, J Michael Cherry, Allan P Davis, Kara Dolinski, Selina S Dwight, Janan T Eppig, et al. 2000. Gene ontology: tool for the unification of biology.Nature genetics25, 1 (2000), 25–29

    2000

  2. [2]

    Junxian Cao, Caifeng Li, Zhao Cui, Shiwen Deng, Tong Lei, Wei Liu, Hongjun Yang, and Peng Chen. 2024. Spatial transcriptomics: a powerful tool in disease understanding and drug discovery.Theranostics14, 7 (2024), 2946

    work page 2024

  3. [3]

    Eduard Chelebian, Christophe Avenel, and Carolina Wählby. 2025. Combining spatial transcriptomics with tissue morphology.Nature Communications16, 1 (2025), 4452

    work page 2025

  4. [4]

    Richard J Chen, Tong Ding, Ming Y Lu, Drew FK Williamson, Guillaume Jaume, Andrew H Song, Bowen Chen, Andrew Zhang, Daniel Shao, Muhammad Shaban, et al. 2024. Towards a general-purpose foundation model for computational pathology.Nature medicine30, 3 (2024), 850–862

    2024

  5. [5]

    Youngmin Chung, Ji Hun Ha, Kyeong Chan Im, and Joo Sang Lee. 2024. Accurate spatial gene expression prediction by integrating multi-resolution features. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. 11591–11600

    work page 2024

  6. [6]

    Haotian Cui, Chloe Wang, Hassaan Maan, Kuan Pang, Fengning Luo, Nan Duan, and Bo Wang. 2024. scGPT: toward building a foundation model for single-cell multi-omics using generative AI.Nature methods21, 8 (2024), 1470–1480

    2024

  7. [7]

    Alexey Dosovitskiy, Lucas Beyer, Alexander Kolesnikov, Dirk Weissenborn, Xi- aohua Zhai, Thomas Unterthiner, Mostafa Dehghani, Matthias Minderer, Georg Heigold, Sylvain Gelly, et al. 2020. An image is worth 16x16 words: Transformers for image recognition at scale.arXiv preprint arXiv:2010.11929(2020)

    Pith/arXiv arXiv 2020

  8. [8]

    Marc Gillespie, Bijay Jassal, Ralf Stephan, Marija Milacic, Karen Rothfels, Andrea Senff-Ribeiro, Johannes Griss, Cristoffer Sevilla, Lisa Matthews, Chuqiao Gong, et al. 2022. The reactome pathway knowledgebase 2022.Nucleic acids research 50, D1 (2022), D687–D692

    work page 2022

  9. [9]

    Zuguang Gu. 2024. simona: a comprehensive R package for semantic similarity analysis on bio-ontologies.BMC genomics25, 1 (2024), 869

    work page 2024

  10. [10]

    Bryan He, Ludvig Bergenstråhle, Linnea Stenbeck, Abubakar Abid, Alma An- dersson, Åke Borg, Jonas Maaskola, Joakim Lundeberg, and James Zou. 2020. Integrating spatial gene expression and breast tumour morphology via deep learning.Nature biomedical engineering4, 8 (2020), 827–834

    work page 2020

  11. [11]

    Jonathan Ho, Ajay Jain, and Pieter Abbeel. 2020. Denoising diffusion probabilistic models.Advances in neural information processing systems33 (2020), 6840–6851

    2020

  12. [12]

    Gao Huang, Zhuang Liu, Laurens Van Der Maaten, and Kilian Q Weinberger

  13. [13]

    Tinglin Huang, Tianyu Liu, Mehrtash Babadi, Wengong Jin, and Zhitao Ying

    work page

  14. [14]

    Guillaume Jaume, Paul Doucet, Andrew Song, Ming Yang Lu, Cristina Alma- gro Pérez, Sophia Wagner, Anurag Vaidya, Richard Chen, Drew Williamson, Ahrong Kim, et al . 2024. Hest-1k: A dataset for spatial transcriptomics and histology image analysis.Advances in Neural Information Processing Systems37 (2024), 53798–53833

    work page 2024

  15. [15]

    Yuran Jia, Junliang Liu, Li Chen, Tianyi Zhao, and Yadong Wang. 2024. THItoGene: a deep learning method for predicting spatial transcriptomics from histological images.Briefings in Bioinformatics25, 1 (2024), bbad464

    work page 2024

  16. [16]

    Liis Kolberg, Uku Raudvere, Ivan Kuzmin, Priit Adler, Jaak Vilo, and Hedi Peterson

    work page

  17. [17]

    Maxim V Kuleshov, Matthew R Jones, Andrew D Rouillard, Nicolas F Fernandez, Qiaonan Duan, Zichen Wang, Simon Koplev, Sherry L Jenkins, Kathleen M Jagodnik, Alexander Lachmann, et al. 2016. Enrichr: a comprehensive gene set enrichment analysis web server 2016 update.Nucleic acids research44, W1 (2016), W90–W97

    work page 2016

  18. [18]

    Sabrina M Lewis, Marie-Liesse Asselin-Labat, Quan Nguyen, Jean Berthelet, Xiao Tan, Verena C Wimmer, Delphine Merino, Kelly L Rogers, and Shalin H Naik

    work page

  19. [19]

    Xinmin Li and Cun-Yu Wang. 2021. From bulk, single-cell to spatial RNA se- quencing.International journal of oral science13, 1 (2021), 36

    work page 2021

  20. [20]

    Kun Liang, Chuanlong Du, Hankun You, and Dan Nettleton. 2018. A hidden Markov tree model for testing multiple hypotheses corresponding to Gene On- tology gene sets.BMC bioinformatics19, 1 (2018), 107

    work page 2018

  21. [21]

    Yuxiang Lin, Ling Luo, Ying Chen, Xushi Zhang, Zihui Wang, Wenxian Yang, Mengsha Tong, and Rongshan Yu. 2024. St-align: A multimodal foundation model for image-gene alignment in spatial transcriptomics.arXiv preprint arXiv:2411.16793(2024)

    Pith/arXiv arXiv 2024

  22. [22]

    Yaron Lipman, Ricky TQ Chen, Heli Ben-Hamu, Maximilian Nickel, and Matt Le

  23. [23]

    Cancer Genome Atlas Network et al. 2015. Genomic classification of cutaneous melanoma.Cell161, 7 (2015), 1681

    work page 2015

  24. [24]

    Hai Dang Nguyen, Nguyen Dang Huy Pham, The Minh Duc Nguyen, Dac Thai Nguyen, Hang Thi Nguyen, and Duong M Nguyen. 2025. MMAP: A Multi- Magnification and Prototype-Aware Architecture for Predicting Spatial Gene Ex- pression. InPacific Rim International Conference on Artificial Intelligence. Springer, 381–395

    work page 2025

  25. [25]

    Steven J Ohsie, G Peter Sarantopoulos, Alistair J Cochran, and Scott W Binder

    work page

  26. [26]

    Minxing Pang, Kenong Su, and Mingyao Li. 2021. Leveraging information in spatial transcriptomics to predict super-resolution gene expression from histology images in tumors.BioRxiv(2021), 2021–11

    work page 2021

  27. [27]

    William Peebles and Saining Xie. 2023. Scalable diffusion models with transform- ers. InProceedings of the IEEE/CVF international conference on computer vision. 4195–4205

    2023

  28. [28]

    Omri Puny, Matan Atzmon, Heli Ben-Hamu, Ishan Misra, Aditya Grover, Edward J Smith, and Yaron Lipman. 2021. Frame averaging for invariant and equivariant network design.arXiv preprint arXiv:2110.03336(2021)

    Pith/arXiv arXiv 2021

  29. [29]

    Sara Sabour, Nicholas Frosst, and Geoffrey E Hinton. 2017. Dynamic routing between capsules.Advances in neural information processing systems30 (2017)

    work page 2017

  30. [30]

    Franco Scarselli, Marco Gori, Ah Chung Tsoi, Markus Hagenbuchner, and Gabriele Monfardini. 2008. The graph neural network model.IEEE transactions on neural networks20, 1 (2008), 61–80

    2008

  31. [31]

    Patrik L Ståhl, Fredrik Salmén, Sanja Vickovic, Anna Lundmark, José Fernández Navarro, Jens Magnusson, Stefania Giacomello, Michaela Asp, Jakub O Westholm, Mikael Huss, et al. 2016. Visualization and analysis of gene expression in tissue sections by spatial transcriptomics.Science353, 6294 (2016), 78–82

    2016

  32. [32]

    Damian Szklarczyk, Katerina Nastou, Mikaela Koutrouli, Rebecca Kirsch, Farrokh Mehryary, Radja Hachilif, Dewei Hu, Matteo E Peluso, Qingyao Huang, Tao Fang, et al. 2025. The STRING database in 2025: protein networks with directionality of regulation.Nucleic acids research53, D1 (2025), D730–D737

    work page 2025

  33. [33]

    Yaqi Tang, Simon Durand, Stéphane Dalle, and Julie Caramel. 2020. EMT-Inducing Transcription Factors, Drivers of Melanoma Phenotype Switching, and Resistance to Treatment.Cancers12, 8 (2020), 2154

    work page 2020

  34. [34]

    Asher Trockman and J Zico Kolter. 2022. Patches are all you need?arXiv preprint arXiv:2201.09792(2022)

    Pith/arXiv arXiv 2022

  35. [35]

    Sipko Van Dam, Urmo Vosa, Adriaan Van der Graaf, Lude Franke, and Joao Pedro de Magalhaes. 2018. Gene co-expression analysis for functional classification and gene–disease predictions.Briefings in bioinformatics19, 4 (2018), 575–592

    work page 2018

  36. [36]

    Ashish Vaswani, Noam Shazeer, Niki Parmar, Jakob Uszkoreit, Llion Jones, Aidan N Gomez, Łukasz Kaiser, and Illia Polosukhin. 2017. Attention is all you need.Advances in neural information processing systems30 (2017)

    2017

  37. [37]

    Petar Veličković, Guillem Cucurull, Arantxa Casanova, Adriana Romero, Pietro Lio, and Yoshua Bengio. 2017. Graph attention networks.arXiv preprint arXiv:1710.10903(2017)

    Pith/arXiv arXiv 2017

  38. [38]

    Chuhan Wang, Adam S Chan, Xiaohang Fu, Shila Ghazanfar, Jinman Kim, Ellis Patrick, and Jean YH Yang. 2025. Benchmarking the translational potential of spatial gene expression prediction from histology.Nature Communications16, 1 (2025), 1544

    work page 2025

  39. [39]

    Hongyi Wang, Xiuju Du, Jing Liu, Shuyi Ouyang, Yen-Wei Chen, and Lanfen Lin. 2025. M2ost: Many-to-one regression for predicting spatial transcriptomics from digital pathology images. InProceedings of the AAAI Conference on Artificial Intelligence, Vol. 39. 7709–7717

    work page 2025

  40. [40]

    Cameron G Williams, Hyun Jae Lee, Takahiro Asatsuma, Roser Vento-Tormo, and Ashraful Haque. 2022. An introduction to spatial transcriptomics for biomedical research.Genome medicine14, 1 (2022), 68

    work page 2022

  41. [41]

    Ronald Xie, Kuan Pang, Sai Chung, Catia Perciani, Sonya MacParland, Bo Wang, and Gary Bader. 2023. Spatially resolved gene expression prediction from histol- ogy images via bi-modal contrastive learning.Advances in Neural Information Processing Systems36 (2023), 70626–70637

    work page 2023

  42. [42]

    Yan Yang, Md Zakir Hossain, Eric A Stone, and Shafin Rahman. 2023. Exemplar guided deep neural network for spatial transcriptomics analysis of gene expres- sion prediction. InProceedings of the IEEE/CVF Winter Conference on Applications of Computer Vision. 5039–5048

    work page 2023

  43. [43]

    Yuansong Zeng, Zhuoyi Wei, Weijiang Yu, Rui Yin, Yuchen Yuan, Bingling Li, Zhonghui Tang, Yutong Lu, and Yuedong Yang. 2022. Spatial transcriptomics prediction from histology jointly through transformer and graph neural networks. Briefings in Bioinformatics23, 5 (2022), bbac297

    work page 2022

  44. [44]

    chylomicron remodeling

    Sichen Zhu, Yuchen Zhu, Molei Tao, and Peng Qiu. 2025. Diffusion generative modeling for spatially resolved gene expression inference from histology images. arXiv preprint arXiv:2501.15598(2025). MM ’26, November 10–14, 2026, Rio de Janeiro, Brazil Zhiwen Xu et al. A Full Main Results Table 5 summarizes the statistics of the nine spatial transcriptomics b...

    Pith/arXiv arXiv 2025

  45. [2008]

    Immunohistochemical characteristics of melanoma.Journal of cutaneous pathology35, 5 (2008), 433–444

    work page 2008

  46. [2017]

    InProceedings of the IEEE conference on computer vision and pattern recognition

    Densely connected convolutional networks. InProceedings of the IEEE conference on computer vision and pattern recognition. 4700–4708

  47. [2021]

    Spatial omics and multiplexed imaging to explore cancer biology.Nature methods18, 9 (2021), 997–1012

    work page 2021

  48. [2022]

    Flow matching for generative modeling.arXiv preprint arXiv:2210.02747 (2022)

    Pith/arXiv arXiv 2022

  49. [2023]

    g: Profiler—interoperable web service for functional enrichment analysis and gene identifier mapping (2023 update).Nucleic acids research51, W1 (2023), W207–W212

    work page 2023

  50. [2025]

    InInternational Conference on Machine Learning

    Scalable Generation of Spatial Transcriptomics from Histology Images via Whole-Slide Flow Matching. InInternational Conference on Machine Learning. PMLR, 25550–25565

    work page

This paper was first reviewed by deepseek-v4-flash on August 5, 2026.