X2Graph for Cancer Subtyping Prediction on Biological Tabular Data

Aparajita Haldar; Mohamed Suliman; Mohammed Amer; Serban Georgescu; Tu Bui

REVIEW 4 major objections 5 minor 39 references

X2Graph for Cancer Subtyping Prediction on Biological Tabular Data

T0 review · 4 major / 5 minor · reviewed 2026-08-07 · deepseek-v4-flash

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

Pith's one-line read Converting each table row into a graph whose edges come from external gene knowledge lets a graph neural network outperform tree-based and deep-learning baselines on three cancer subtyping datasets.

desk verdict X2Graph is a genuinely new way to inject biological knowledge into tabular GNNs, but the paper never rules out that its gains come from the knowledge-base feature subset alone. read the letter →

arxiv 2505.23334 v1 pith:TISMQNT4 submitted 2025-05-29 cs.LG

Tu Bui , Mohamed Suliman , Aparajita Haldar , Mohammed Amer , Serban Georgescu This is my paper

classification cs.LG

keywords tabulardatagraphneuralnetworksinductivebiasknowledgebasecancersubtypingTCGABRCAsmall-samplelearningbiological

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

X2Graph claims that converting each row of a small biological table into a graph, with edges defined by external knowledge bases such as gene interaction networks, lets a standard graph neural network beat tree-based and deep-learning baselines on cancer subtyping. The method is tested on three TCGA BRCA datasets (copy number variation, RNA expression, and clinical records) with hundreds of samples and tens of thousands of features. The reported gains come from the inductive bias that only known-related features talk to each other, reducing overfitting in the high-dimensional, low-sample regime.

What carries the argument

The central mechanism is the row-to-graph conversion: a node vector encodes the pair (feature index, cell value), which makes the graph view-invariant and lets the original row be reconstructed from the graph; an edge is present when the knowledge base's relation matrix links the two features. For copy-number-variation data, nodes with zero values are pruned because they indicate no copy-number change, further strengthening the inductive bias. For two of the datasets, predictions from three separate knowledge-base graphs are combined by a learned weighted sum.

What would settle it

Train X2Graph on the clinical dataset with the knowledge-base edges randomly rewired while preserving node degrees; if the model still matches or surpasses the strongest tree-based baseline, the paper's inductive-bias explanation would be unsupported. Alternatively, remove a clinically documented edge, such as the link between the HER2 immunohistochemistry score and HER2 fluorescence in situ hybridization status, and check whether AUC drops toward the unconstrained deep-learning baseline.

Watch

Extended reading notes

Core claim

The paper proposes X2Graph, a graph-based deep learning method for biological tabular data. Each sample row is converted into a graph where nodes are the cell values paired with the feature index, and edges come from an external knowledge base that connects related features. Message passing is then applied with standard graph-neural-network layers. On the three TCGA BRCA datasets, X2Graph achieves the highest accuracy, macro AUC, and Cohen's kappa among all baselines, with F1 second to the strongest tree-based model on clinical data. The paper attributes the gains to the inductive bias supplied by the knowledge base.

Load-bearing premise

The method's gain rests on the assumption that the external knowledge bases already contain the most crucial relations between features for the prediction task; if a key relation is missing or an edge is spurious, the graph's inductive bias can mislead the model.

Editorial extensions

If this is right

On the copy-number-variation and RNA datasets, the fused X2Graph model outperforms the strongest tree-based baseline on all four reported metrics.
The ablation study shows that including the feature index in node vectors is essential, improving AUC by at least 15% regardless of node pruning.
Late fusion of three knowledge-base graphs outperforms an early-fusion multi-edge version by about one point of AUC on the copy-number-variation data.
The interpretability analysis identifies top contributing features, and a larger fraction of these features have literature evidence for breast-cancer linkage than for the tree-based and deep-learning baselines.

Reading between the lines

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

The same row-to-graph recipe may transfer to any tabular domain, medical or not, where a reliable graph of column relationships exists; a natural next experiment is to apply it to financial or sensor data with known dependencies.
The paper's claim predicts a specific degradation pattern: randomly rewiring knowledge-base edges while preserving node degree should erode performance back toward the unconstrained deep-learning baseline.
The manually built clinical knowledge base (about fifty nodes and two hundred edges) suggests that even small hand-curated relation graphs can steer deep learning, which could make the approach usable in domains lacking large public knowledge bases.

Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

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

Desk Editor's Note

read the letter

Worth a look, but the headline claim has a hole. X2Graph converts each row into a graph with edges from external KBs. That is genuinely different from IGGNet/TABGLM, which infer edges from the data, and unlike Plato it does not require a huge KB. The write-up is clear, the ablation of the gene-ID positional encoding is a real contribution, and the interpretability check against the literature is a nice verification step.

The soft spot is the missing control. Algorithm 1 starts with S = S_D ∩ S_B, so for CNV and RNA the model sees roughly 4,152 MoNDO genes instead of 20k+ columns, and all baselines are trained on the full column set. The paper credits the KB's inductive bias for the gains, but nothing rules out that simply dropping the unannotated columns is what helps. A plain MLP or XGBoost on the same S would settle it. The stress-test note is right, and this is a correctness risk to the attribution, not just a completeness worry. The Clinical KB also selects 50 of 138 columns, so the same issue applies there.

The quantitative issues are smaller but real: no error bars or significance tests, and the margins over XGBoost are thin (CNV +0.010 accuracy, RNA +0.005). On Clinical, XGBoost actually has the higher F1, so 'superior performance across three datasets' overstates the case. The unpublished, anonymous baseline [15] is used as the MLP comparator and as the source of the fusion idea, which adds opacity to the comparison.

The method is plausible, the experiments are extensive, and the direction is worth pursuing for medical ML and tabular GNN researchers. But the central mechanism claim is unproven until we see the same-feature-subset baselines. A serious referee should ask for that, plus confidence intervals and release of the handcrafted clinical KB and code. I would send it out rather than desk-reject: it is a solid workshop-to-conference submission with one clear empirical gap.

Referee Report

4 major / 5 minor

Summary. The paper proposes X2Graph, a method that converts each row of a biological tabular dataset into a graph whose nodes are (feature index, cell value) pairs and whose edges are defined by external knowledge bases (MoNDO, Reactome, BioGrid, and a manually constructed Clinical KB). Node representations are learned by a GNN, and predictions from KB-specific models are combined with a late-fusion layer. The method is evaluated on TCGA BRCA CNV, RNA, and Clinical data using 10-fold cross-validation, comparing accuracy, AUC, F1, and Cohen's kappa against MLP, TabPFN, XGBoost, LASSO, and Ridge. The authors report state-of-the-art results on most metrics, and also present a GNNExplainer-based feature-importance analysis plus ablations of node pruning and feature-index encoding.

Significance. If the performance attribution were established, X2Graph would be a useful contribution: it injects domain knowledge into tabular deep learning without requiring a large auxiliary knowledge graph, and the row-to-graph conversion with feature-index coding is simple, interpretable, and clearly described. The use of public TCGA data and three distinct gene-interaction KBs is a strength, as is the explicit ablation of ID indexing and node pruning. However, the central mechanistic claim — that the KB-defined edges, rather than the KB-induced feature subset, drive the gains — is not tested, and the statistical support for the headline results is weak. The paper gives no code, and the MLP baseline relies on an unpublished 'under review' reference. With a same-subset baseline comparison and variance/significance reporting, the claims could become solid; in its current form the evidence is suggestive rather than conclusive.

major comments (4)

[§IV-A (Algorithm 1), §V-C, Tables II-III] Algorithm 1 restricts the feature set to S = S_D ∩ S_B, so X2Graph trains only on the KB-intersection features. For CNV the full table has 23,286 columns while the MoNDO KB has 4,152 nodes, and for RNA the numbers are 20,530 versus 4,152; the baselines in Tables II and III are trained on the full feature set. The reported gains over XGBoost (e.g., +0.010 accuracy on CNV, +0.005 on RNA) could therefore be due entirely to KB-guided feature subsetting. Section V-C attributes the gains to 'the strong inductive bias introduced by the KBs', but Table V ablates only node pruning and ID indexing on the MoNDO graph and never evaluates an MLP, XGBoost, or LASSO on the same feature subset S. Please add a same-subset baseline comparison, and ideally a control with random edges or a feature-count-matched model, to separate feature selection from message passing.
[§V-C, Tables II-IV] No measure of variance is reported: Tables II-IV give only point estimates from 10-fold cross-validation. Many differences are small (CNV accuracy 0.7410 vs 0.7312 for XGBoost; RNA accuracy 0.9242 vs 0.9194), and on Clinical data XGBoost's F1 (0.6800) exceeds X2Graph's (0.6523), so the abstract's claim of 'superior performance' across the three datasets is not accurate. Without standard deviations, confidence intervals, or a paired significance test across the folds, the state-of-the-art claim is under-supported. Please report fold-level variability, run a paired test (e.g., Wilcoxon or corrected t-test), and soften the conclusions accordingly.
[§III-B, §V-D, §VI] The Clinical KB is manually constructed from literature that is also used to interpret the Clinical features, and §V-D states that Clinical interpretability results are omitted because 'all features have a clear correlation with cancer'. This is an acknowledged limitation, but it has consequences for the paper's third contribution: for Clinical data the 'external knowledge' is not independent of the outcome, so the edge-structure inductive bias cannot be credited with discovering relevant features. The interpretability conclusions should be scoped to CNV/RNA, and the manual Clinical KB should be described in enough detail (nodes, edges, and evidence per edge) to allow readers to assess potential label leakage into the graph definition.
[§II and §V-C] The paper motivates X2Graph against other GNN-based tabular methods (IGGNet, TABGLM, Plato in Section II), but the experiments compare only against MLP, TabPFN, XGBoost, LASSO, and Ridge. The claim of state-of-the-art performance is therefore limited to non-graph baselines; to substantiate the advantage of the graph conversion, the authors should include a GNN-based tabular baseline (e.g., IGGNet or a GNN on a fully connected graph) in the comparison, or explicitly restrict the claim to the evaluated baseline families.

minor comments (5)

[§III-A] The text says 'Electrical Heath Records'; this should be 'Electronic Health Records'.
[§IV-A, Algorithm 1] The function nodeFeat(v, s, **kwargs) is never defined; please specify how the node feature vector is assembled for each modality and what the allowed kwargs are.
[§V-D, Fig. 4(d)] The 'fraction of top genes with links to breast cancer' comparison does not state how many top genes were selected for each method or how the baseline attributions (split counts for XGBoost, integrated gradients for MLP) were normalized; please make this explicit.
[§V-B] The hyperparameter search ('up to 50 random search iterations') is not accompanied by search ranges or the selected hyperparameters for the baselines; this information is needed to rule out undertuned baselines and to make the comparison reproducible.
[References and §V-C] Reference [15] is an anonymous 'under review' work used both as the MLP baseline and as inspiration for the fusion method; since the reader cannot verify this baseline, please use a published baseline or provide a detailed implementation description, and consider moving the citation to a public version if one exists.

Circularity Check

0 steps flagged · score 2.0 of 10

No circular derivation: X2Graph is evaluated on held-out folds against external baselines, and the KB edges are not fitted to the labels; the only minor self-citation is [15], which is not load-bearing for the core claim.

full rationale

The central claim is that converting tabular rows to graphs with knowledge-base-defined edges improves cancer subtyping. The KBs (MoNDO, Reactome, BioGrid, Clinical) are constructed from external literature and are not fitted to the TCGA labels, so the message-passing outputs are not equal to any input by construction. Algorithm 1 does intersect the dataset features with KB features (S = S_D ∩ S_B), meaning X2Graph trains on a restricted feature set, but this is a feature-selection property of the method and not a fitted-input-called-prediction; the same restriction is applied consistently to all X2Graph variants, and the ablations in Table V vary graph components. The main evaluation is 10-fold cross-validation against public baselines (XGBoost, TabPFN, LASSO, Ridge) and external published deep-learning methods, so the performance result has independent content. The only mild self-citation is reference [15], an anonymous 'under review' manuscript used for patient filtering, as the MLP baseline, and as fusion inspiration; it is not load-bearing because XGBoost and TabPFN are external and the core KB-versus-overfitting argument does not rest on [15]. The manually constructed Clinical KB is a stated limitation and its assumption is explicit, but this is an assumption about domain knowledge, not a circular reduction. The missing same-subset baseline (MLP or XGBoost on S) is a correctness and attribution risk, not a circularity, and is noted here for consideration elsewhere.

Assumptions & free parameters 4 free parameters · 5 assumptions · 0 invented entities

The central claim rests on tuning hyperparameters, on the completeness and correctness of external KBs (including an unreleased handcrafted Clinical KB), and on the assumption that zero-valued CNV genes are uninformative. None of these are fitted to the labels through circular equations, but they are assumptions the reader must grant.

free parameters (4)

Feature embedding dimension d per KB = 128 to 1024 (Table I)
Chosen by random search; affects model capacity and performance.
Graph architecture per KB = SG, GEN, GCN, SAGE, GAT, PNA (Table I)
Selected by random search; part of the model design.
Random node/edge sampling budget = 6400 nodes, 3200 edges for CNV/RNA; none for Clinical
Hand-picked data augmentation choice; affects training and results.
Learning rate and training epochs = 1e-3 or 1e-4 learning rate; 200 or 400 epochs
Tuned standard training hyperparameters.

assumptions (5)

standard math Backpropagation and standard GNN message-passing operations are assumed to work as expected.
Used throughout Section IV-B without proof.
domain assumption MoNDO, Reactome, and BioGrid edges represent associations relevant to breast cancer subtyping.
Section III-B: edges connect genes with shared disease associations, pathways, or physical/genetic interactions.
domain assumption The manually constructed Clinical KB contains the crucial feature relations for subtyping.
Section III-B and Limitations: 'our assumption is that the most crucial relations between features are already defined in these KBs.'
domain assumption Zero-valued CNV genes are not informative for cancer subtyping.
Section IV-A node pruning: 'genes with the value of zero represent no variants ... they are usually not important for cancer subtyping.'
domain assumption Gene ID indexing makes graphs permutation invariant and allows reconstruction of the original row.
Section IV-A: 'this makes the mapping between table row and output graph bidirectional.'

how reviews work

0 comments

Cite this review

Pith. "Pith review of X2Graph for Cancer Subtyping Prediction on Biological Tabular Data." pith.science (2026). https://pith.science/paper/TISMQNT4

@misc{pith2026250523334,
  author       = {Pith},
  title        = {Pith review of: X2Graph for Cancer Subtyping Prediction on Biological Tabular Data},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/TISMQNT4}},
  note         = {Machine review of arXiv:2505.23334}
}

read the original abstract

Despite the transformative impact of deep learning on text, audio, and image datasets, its dominance in tabular data, especially in the medical domain where data are often scarce, remains less clear. In this paper, we propose X2Graph, a novel deep learning method that achieves strong performance on small biological tabular datasets. X2Graph leverages external knowledge about the relationships between table columns, such as gene interactions, to convert each sample into a graph structure. This transformation enables the application of standard message passing algorithms for graph modeling. Our X2Graph method demonstrates superior performance compared to existing tree-based and deep learning methods across three cancer subtyping datasets.

Figures

Figures reproduced from arXiv: 2505.23334 by the authors.

**Figure 1.** X2Graph converts each table row into a graph. The cell values become node features, while the edge connection comes from the KB. The (.) [PITH_FULL_IMAGE:figures/full_fig_p002_1.png] view at source ↗

**Figure 2.** Subgraph visualization for three gene KBs. Each subgraph shows 1-hop neighbor connections centered at gene BRCA1. [PITH_FULL_IMAGE:figures/full_fig_p003_2.png] view at source ↗

**Figure 3.** (Top) PR curves and Average Precision (AP) of X2Graph and baselines on the three benchmarks across the 10-fold cross-validation test sets. [PITH_FULL_IMAGE:figures/full_fig_p004_3.png] view at source ↗

Figures from the paper (2 more)

**Figure 4.** Figure 4: (a-c) Top k most importance features identified from X2Graph models for CNV, RNA and Clinical data. (d) Fractions of these features evidenced [PITH_FULL_IMAGE:figures/full_fig_p004_4.png]

**Figure 5.** Figure 5: Contributions of each classes and KB in multigraph fusion for (a) CNV and (b) RNA, averaged across 10-fold cross validation. [PITH_FULL_IMAGE:figures/full_fig_p006_5.png]

Discussion (0). Sign in to comment.

Reference graph

Works this paper leans on

39 extracted references · 33 canonical work pages

[15]

Scalable and loosely-coupled multimodal deep learn- ing for breast cancer subtyping,

A. Anonymous, “Scalable and loosely-coupled multimodal deep learn- ing for breast cancer subtyping,” 2025, under review

work page 2025
[1]

The future of ai-enabled health: Lead- ing the way,

World Economic Forum, “The future of ai-enabled health: Lead- ing the way,” https://reports.weforum.org/docs/WEF The Future of AI Enabled Health 2025.pdf, 2025, accessed: 2025-01-27

work page 2025
[2]

Why do tree-based models still outperform deep learning on typical tabular data?

L. Grinsztajnet al., “Why do tree-based models still outperform deep learning on typical tabular data?”NeurIPS, vol. 35, pp. 507–520, 2022

work page 2022
[3]

Position: Why tabular foundation models should be a research priority,

B. Van Breugel and M. Van Der Schaar, “Position: Why tabular foundation models should be a research priority,” inICML, ser. PMLR, vol. 235, 21–27 Jul 2024, pp. 48 976–48 993. [Online]. Available: https://proceedings.mlr.press/v235/van-breugel24a.html

work page 2024
[4]

A guide to deep learning in healthcare,

Estevaet al., “A guide to deep learning in healthcare,”Nature medicine, vol. 25, no. 1, pp. 24–29, 2019

work page 2019
[5]

Implications of gene copy-number variation in health and diseases,

S. H. Almal and H. Padh, “Implications of gene copy-number variation in health and diseases,”Journal of human genetics, vol. 57, no. 1, pp. 6–13, 2012

work page 2012
[6]

Transcriptomic landscape of breast cancers through mrna sequencing,

Eswaranet al., “Transcriptomic landscape of breast cancers through mrna sequencing,”Scientific reports, vol. 2, no. 1, p. 264, 2012

work page 2012
[7]

Rna-seq analyses generate comprehensive transcrip- tomic landscape and reveal complex transcript patterns in hepatocel- lular carcinoma,

Huanget al., “Rna-seq analyses generate comprehensive transcrip- tomic landscape and reveal complex transcript patterns in hepatocel- lular carcinoma,”PloS one, vol. 6, no. 10, p. e26168, 2011

work page 2011

Show all 39 references

[8]

Artificial intelligence in healthcare: past, present and future,

F. Jianget al., “Artificial intelligence in healthcare: past, present and future,”Stroke and vascular neurology, vol. 2, no. 4, 2017

2017
[9]

Regression shrinkage and selection via the lasso,

R. Tibshirani, “Regression shrinkage and selection via the lasso,” Royal Stat. Society, vol. 58, no. 1, pp. 267–288, 1996

1996
[10]

Ridge regression in practice,

D. W. Marquardt and R. D. Snee, “Ridge regression in practice,”The American Statistician, vol. 29, no. 1, pp. 3–20, 1975

1975
[11]

Random forests,

L. Breiman, “Random forests,”Machine learning, vol. 45, pp. 5–32, 2001

2001
[12]

Xgboost: A scalable tree boosting system,

T. Chen and C. Guestrin, “Xgboost: A scalable tree boosting system,” inProc. ACM SIGKDD, 2016, pp. 785–794

2016
[13]

Lung cancer subtype diagnosis by fusing image- genomics data and hybrid deep networks,

X. Wanget al., “Lung cancer subtype diagnosis by fusing image- genomics data and hybrid deep networks,”IEEE/ACM Trans. Comp. biology and bioinformatics, vol. 20, no. 1, pp. 512–523, 2021

2021
[14]

A hybrid deep learning model for predicting molecular subtypes of human breast cancer using multimodal data,

T. Liuet al., “A hybrid deep learning model for predicting molecular subtypes of human breast cancer using multimodal data,”Irbm, vol. 43, no. 1, pp. 62–74, 2022

2022
[16]

Tabpfn: A transformer that solves small tabular classification problems in a second,

N. Hollmannet al., “Tabpfn: A transformer that solves small tabular classification problems in a second,”arXiv preprint arXiv:2207.01848, 2022

2022 arXiv
[17]

Tabnet: Attentive interpretable tabular learning,

S. ¨O. Arik and T. Pfister, “Tabnet: Attentive interpretable tabular learning,” inAAAI, vol. 35, 2021, pp. 6679–6687

2021
[18]

Tabtransformer: Tabular data modeling using contextual embeddings,

X. Huang, A. Khetan, M. Cvitkovic, and Z. Karnin, “Tabtransformer: Tabular data modeling using contextual embeddings,”arXiv preprint arXiv:2012.06678, 2020

2012 arXiv
[19]

Tabular data: Deep learning is not all you need,

R. Shwartz-Ziv and A. Armon, “Tabular data: Deep learning is not all you need,”Information Fusion, vol. 81, pp. 84–90, 2022

2022
[20]

Deep neural networks for high dimension, low sample size data

B. Liuet al., “Deep neural networks for high dimension, low sample size data.” inIJCAI, vol. 2017, 2017, pp. 2287–2293

2017
[21]

Spurious correlations in machine learning: A survey,

W. Yeet al., “Spurious correlations in machine learning: A survey,” arXiv preprint arXiv:2402.12715, 2024

2024
[22]

Vime: Extending the success of self-and semi- supervised learning to tabular domain,

J. Yoonet al., “Vime: Extending the success of self-and semi- supervised learning to tabular domain,”NeurIPS, vol. 33, pp. 11 033– 11 043, 2020

2020
[23]

Graph neural networks for tabular data learning,

C.-T. Liet al., “Graph neural networks for tabular data learning,” in ICDE. IEEE, 2023, pp. 3589–3592

2023
[24]

Interpretable graph neural networks for tabular data,

A. Alkhatibet al., “Interpretable graph neural networks for tabular data,” inECAI 2024. IOS Press, 2024, pp. 1848–1855

2024
[25]

Tabglm: Tabular graph language model for learning transferable representations through multi-modal consistency mini- mization,

A. Majeeet al., “Tabglm: Tabular graph language model for learning transferable representations through multi-modal consistency mini- mization,” inProc. AAAI, vol. 39, no. 18, 2025, pp. 19 387–19 395

2025
[26]

High dimensional, tabular deep learning with an auxiliary knowledge graph,

C. Ruizet al., “High dimensional, tabular deep learning with an auxiliary knowledge graph,”NeurIPS, vol. 36, 2024

2024
[27]

The monarch initiative: an integrative data and analytic platform connecting phenotypes to genotypes across species,

Mungallet al., “The monarch initiative: an integrative data and analytic platform connecting phenotypes to genotypes across species,”Nucleic acids research, vol. 45, no. D1, pp. D712–D722, 2017

2017
[28]

The reactome pathway knowledgebase,

Fabregatet al., “The reactome pathway knowledgebase,”Nucleic acids research, vol. 46, no. D1, pp. D649–D655, 2018

2018
[29]

The biogrid interaction database: 2019 update,

Oughtredet al., “The biogrid interaction database: 2019 update,” Nucleic acids research, vol. 47, no. D1, pp. D529–D541, 2019

2019
[30]

The multilayer community structure of medulloblastoma,

I. N ´u˜nez-Carpinteroet al., “The multilayer community structure of medulloblastoma,”Iscience, vol. 24, no. 4, 2021

2021
[31]

Concordance of her2 immunohistochemistry and fluorescence in situ hybridization using tissue microarray in breast cancer,

D. Furreret al., “Concordance of her2 immunohistochemistry and fluorescence in situ hybridization using tissue microarray in breast cancer,”Anticancer Research, vol. 37, no. 6, pp. 3323–3329, 2017

2017
[32]

Attention is all you need,

A. Vaswaniet al., “Attention is all you need,”NeurIPS, 2017

2017
[33]

Gnnexplainer: Generating explanations for graph neural networks,

Z. Yinget al., “Gnnexplainer: Generating explanations for graph neural networks,”NeurIPS, vol. 32, 2019

2019
[34]

Uniprot: the universal protein knowledgebase in 2025,

T. U. Consortium, “Uniprot: the universal protein knowledgebase in 2025,”Nucleic Acids Research, vol. 53, no. D1, pp. D609–D617, 2025

2025
[35]

The oncogenic roles and clinical implications of yap/taz in breast cancer,

J. Luoet al., “The oncogenic roles and clinical implications of yap/taz in breast cancer,”British Journal of Cancer, vol. 128, no. 9, pp. 1611– 1624, 2023

2023
[36]

Glucose-6-phosphate dehydrogenase and trans- ketolase: Key factors in breast cancer progression and therapy,

X. Zhenet al., “Glucose-6-phosphate dehydrogenase and trans- ketolase: Key factors in breast cancer progression and therapy,” Biomedicine & Pharmacotherapy, vol. 176, p. 116935, 2024

2024
[37]

Adam10: a new player in breast cancer progres- sion?

M. Mulloolyet al., “Adam10: a new player in breast cancer progres- sion?”British journal of cancer, vol. 113, no. 6, pp. 945–951, 2015

2015
[38]

Estrogen and progesterone receptor testing in breast cancer: Asco/cap guideline update,

K. H. Allisonet al., “Estrogen and progesterone receptor testing in breast cancer: Asco/cap guideline update,”Journal of Clinical Oncology, vol. 38, no. 12, pp. 1346–1366, 2020

2020
[39]

Axiomatic attribution for deep networks,

M. Sundararajan, A. Taly, and Q. Yan, “Axiomatic attribution for deep networks,” inICML. PMLR, 2017, pp. 3319–3328

2017

Pith tools

Reviewed August 7, 2026 · model on record in the stance chip above.

[1] [15]

Scalable and loosely-coupled multimodal deep learn- ing for breast cancer subtyping,

A. Anonymous, “Scalable and loosely-coupled multimodal deep learn- ing for breast cancer subtyping,” 2025, under review

work page 2025

[2] [1]

The future of ai-enabled health: Lead- ing the way,

World Economic Forum, “The future of ai-enabled health: Lead- ing the way,” https://reports.weforum.org/docs/WEF The Future of AI Enabled Health 2025.pdf, 2025, accessed: 2025-01-27

work page 2025

[3] [2]

Why do tree-based models still outperform deep learning on typical tabular data?

L. Grinsztajnet al., “Why do tree-based models still outperform deep learning on typical tabular data?”NeurIPS, vol. 35, pp. 507–520, 2022

work page 2022

[4] [3]

Position: Why tabular foundation models should be a research priority,

B. Van Breugel and M. Van Der Schaar, “Position: Why tabular foundation models should be a research priority,” inICML, ser. PMLR, vol. 235, 21–27 Jul 2024, pp. 48 976–48 993. [Online]. Available: https://proceedings.mlr.press/v235/van-breugel24a.html

work page 2024

[5] [4]

A guide to deep learning in healthcare,

Estevaet al., “A guide to deep learning in healthcare,”Nature medicine, vol. 25, no. 1, pp. 24–29, 2019

work page 2019

[6] [5]

Implications of gene copy-number variation in health and diseases,

S. H. Almal and H. Padh, “Implications of gene copy-number variation in health and diseases,”Journal of human genetics, vol. 57, no. 1, pp. 6–13, 2012

work page 2012

[7] [6]

Transcriptomic landscape of breast cancers through mrna sequencing,

Eswaranet al., “Transcriptomic landscape of breast cancers through mrna sequencing,”Scientific reports, vol. 2, no. 1, p. 264, 2012

work page 2012

[8] [7]

Rna-seq analyses generate comprehensive transcrip- tomic landscape and reveal complex transcript patterns in hepatocel- lular carcinoma,

Huanget al., “Rna-seq analyses generate comprehensive transcrip- tomic landscape and reveal complex transcript patterns in hepatocel- lular carcinoma,”PloS one, vol. 6, no. 10, p. e26168, 2011

work page 2011