Pith. sign in

REVIEW 4 major objections 5 minor 24 references

HyperNetWalk: A Unified Framework for Personalized and Cohort-Level Cancer Driver Gene Identification via Reverse Inference on Layered Signaling-Regulatory Network

T0 review · 4 major / 5 minor · reviewed 2026-08-03 · deepseek-v4-flash

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

Pith's one-line read HyperNetWalk claims that one unsupervised two-stage random-walk pipeline—reverse-tracing transcriptional dysregulation through protein interaction and gene-regulatory networks, then sharing signal across patients via hypergraph walks—ranks

desk verdict Genuinely unified two-level driver-gene framework with honest benchmarking, but the reverse-inference claim lacks a hub-bias control and the two abstracts disagree. read the letter →

arxiv 2601.01337 v2 pith:WWC76AL7 submitted 2026-01-04 q-bio.QM q-bio.MN

Xueqing Xu , Yonghang Gao , Duanchen Sun , Ling-Yun Wu This is my paper · ORCID
classification q-bio.QMq-bio.MN
keywords cancerdrivergenesrandomwalkwithrestarthypergraphdiffusionprotein-proteininteractionnetworkgeneregulatorypersonalizedmedicinecohort-levelpredictionTCGA
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

HyperNetWalk proposes to unify two tasks that are usually separate: finding driver genes for a single patient and finding driver genes for a whole cancer cohort. The central idea is to treat driver identification as reverse inference: instead of asking which mutated genes are most frequent, walk backward from the observed transcriptional disruption of a patient to the mutated genes that could plausibly cause it, using a layered network of protein-protein interactions (upstream signaling), transcription-factor-to-target regulation (downstream control), and tumor/normal co-expression. A first stage performs patient-specific random walks with restart on these induced subnetworks; a second stage builds hypergraphs whose hyperedges are patients and runs random walks that let samples borrow strength from similar samples, producing both personalized rankings and cohort rankings from the same machinery. The paper evaluates this on 12 TCGA cancer types and reports that the method is competitive with DawnRank for personalized predictions, is second only to DriverMP for cohort-level predictions, and best captures cancer-type-specific drivers when checked against IntOGen reference sets. If these results hold, the contribution is a single unsupervised pipeline that gives both a clinician and a researcher a ranked list of candidate drivers without needing labeled training data.

What carries the argument

The load-bearing object is a layered signaling-regulatory network: an undirected PPI network approximating upstream signaling connected to a directed gene regulatory network (transcription factors as interface nodes) approximating downstream transcriptional control, with a dynamic co-expression network layered on top. On this layered representation the paper runs random walks with restart (RWR), a diffusion process in which a walker moves along network edges and periodically resets to a prior distribution; the reverse-inference mechanism is encoded by reversing GRN edges and by using differential-expression z-scores as the restart prior, so probability flows from observed downstream perturba

What would settle it

Rerun HyperNetWalk on the same 12 cohorts with the protein-protein interaction and gene-regulatory subnetworks replaced by degree-preserving random networks, keeping the expression-based restart priors unchanged. If Precision@K against the known-driver list stays nearly the same, the claimed reverse-inference mechanism is not what carries the ranking. As a second check, if ranking mutated genes by network degree and regulatory out-degree alone matches the top-200 recall, then hub bias explains most of the result.

Watch

Extended reading notes

Core claim

HyperNetWalk claims driver identification can be unified at two scales. For each patient, it induces subnetworks from mutated genes: PPI and GRN supply topology (GRN reversed, so walks run upstream), and a co-expression network gives perturbation signal from z-scores. Random walks with restart convert these into scores. Stage II makes hypergraphs with patients as hyperedges and runs hypergraph random walks to diffuse scores across similar patients, producing personalized and cohort rankings. On 12 TCGA cancers, it reports personalized precision on par with DawnRank, cohort performance behind only DriverMP, and clear cancer-specific enrichment against IntOGen; ablations support the design.

Load-bearing premise

The evaluation assumes that the curated list of known cancer driver genes is a complete and unbiased truth label, even though the model is built to reward the same network-centrality properties that helped genes get onto that list.

Editorial extensions

If this is right

  • A single pipeline can supply both patient-specific shortlists and cohort-level driver lists, so clinical and research use cases no longer require separate models.
  • Rare, low-frequency drivers can be recovered if they sit upstream of strongly perturbed downstream genes, rather than being filtered out by mutation-recurrence thresholds.
  • Cancer-type-specific driver sets can be recovered, which is useful for tissue-specific therapeutic targeting and for distinguishing general drivers like TP53 from context-dependent ones.
  • High precision at the top of the ranked list means fewer candidates need experimental validation, making downstream functional screens cheaper.

Reading between the lines

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

  • A decisive test of the reverse-inference story would be ablating the topology terms: if rankings hold when PPI/GRN subnetworks are replaced with degree-matched random graphs and only the expression-perturbation prior remains, then the causal 'reverse walk' narrative is not the active ingredient; the paper's own ablation goes partway but a random-network control would isolate topology from diffusio
  • The hypergraph refinement stage is score-agnostic; feeding it scores from any per-patient ranker (e.g., a simple differential-expression summary) would show whether cross-sample diffusion, not Stage I's specific priors, drives the gain.
  • Because the benchmark uses the curated Cancer Gene Census list, a set itself enriched for well-studied hub genes, the reported precision may overstate performance on undiscovered drivers; checking against held-out recent additions to the census, or against functional screens, would test generalization.
  • The diagonal cancer-type specificity pattern suggests a possible second use the paper does not develop: using the method's cohort-level rankings to assign a tumor sample to its tissue of origin, or to flag predicted drivers whose tissue association is ambiguous.
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

4 major / 5 minor

Summary. The paper proposes HyperNetWalk, an unsupervised two-stage framework for cancer driver gene identification that combines static (PPI, GRN) and dynamic (co-expression) networks with hypergraph diffusion. In Stage I, sample-specific RWR on patient-specific subnetworks yields per-gene scores; in Stage II, hypergraph RWR across similar samples refines these scores for both personalized and cohort-level predictions. The method is benchmarked on 12 TCGA cancer types against eight existing methods, with claims of personalized performance comparable to DawnRank and cohort-level performance second only to DriverMP, plus an IntOGen-based analysis of cancer-type specificity.

Significance. If the reported performance is robust and the mechanism is genuinely 'reverse inference' rather than a proxy for node degree, the paper offers a useful unified framework for both personalized and population-level driver prioritization. The strengths are the wide benchmark (12 cancer types, 8 baselines, multiple metric families), the fully unsupervised architecture, and the explicit integration of multiple network types. However, the absence of a degree-controlled baseline, the lack of statistical testing across datasets, and the unverified ablation claims mean that the central mechanistic claim is not yet established. The paper is likely of interest to the computational cancer genomics community, but the current evidence does not fully separate the method's novelty from known hub properties of driver genes.

major comments (4)
  1. [§2.2, §3.3.1, Eq. (16)] The central concern is hub-bias. The paper's own Fig. S1(a,b) shows that known CGC Tier 1 drivers have significantly higher PPI degree and GRN out-degree than non-drivers. Stage I's RWR (Eqs. 1–2) uses row-normalized transition matrices on the induced PPI/GRN subnetworks, which rewards high-degree nodes, and the final Stage I score is dominated by the static networks because γ = 0.8 in Eq. (16). No degree-only baseline (e.g., ranking by PPI degree) or degree-preserving network null is included in the benchmark. Consequently, the reported Precision@K and pAUPRC values may largely reflect hub detection rather than the claimed reverse-inference mechanism. Please add a control where only topological prominence is used, and/or an ablation that removes the static network component, to demonstrate that the co-expression and hypergraph components contribute beyond degree.
  2. [§2.3, Table S2] Table S2 provides only point estimates, and the text makes claims such as 'consistently outperforms' and 'significantly higher' without any error bars, confidence intervals, or between-method statistical tests. In fact, reading the table, DawnRank has higher AUROC in 10 of 12 cancer types and higher pAUROC in 7 of 12, while DriverMP has higher pAUPRC in most cancer types. The hedged phrasing in the results is appropriate, but the abstract's 'outperformed representative existing methods' is not supported. Please perform paired significance tests across the 12 cancer types (e.g., Wilcoxon signed-rank test) and report which differences are statistically significant.
  3. [§4 (Discussion), data availability] The manuscript states 'Ablation analyses supported the contributions of the reverse inference formulation and layered network architecture,' but no ablation results are presented in the main text or in the listed supplementary figures/tables (Figs. S1–S7, Tables S1–S2). Without these results, the contribution of each component (PPI, GRN, co-expression, hypergraph refinement, ME prior) cannot be assessed, and the claim that the architecture matters is unverified. Please include ablation tables or remove the claim.
  4. [§3.3–§3.5] The method has many free parameters (λ, θ, γ, β, τ, top-500 DEGs, σ, α, the ME cutoff |V_m|≥200, and the personalized horizon N_c), all set to fixed values without sensitivity analysis. Given that the performance differences between HyperNetWalk and the best baselines are often small (a few percent in pAUPRC/pAUROC), the conclusions may depend on these choices. Please provide a sensitivity analysis over the most influential parameters (e.g., λ, γ, τ, σ, α) or otherwise justify the fixed values.
minor comments (5)
  1. [Abstract vs. §2.3] The abstract claims HyperNetWalk 'outperformed representative existing methods,' while the results state that it is 'comparable to DawnRank' and 'second-best behind DriverMP.' Please harmonize the wording.
  2. [Eq. (10)] There is a typo: 'the 10 will not update significantly' should be 'the equation will not update significantly.' Also, the convergence criterion for the RWR iterations is not specified.
  3. [Eq. (1)] The transition formula does not define the convention for nodes with zero degree in one network; a 0/0 term may arise. Please clarify that such terms are treated as 0.
  4. [§3.4.1, Eq. (17)] The neighbor set requires sharing at least two mutated genes with the target sample. Many samples, especially those with low mutation burden, may have no neighbors, effectively reducing the personalized hypergraph to a single hyperedge. Please state how such cases are handled.
  5. [§3.3.2, Eq. (11)] The 'dynamic co-expression network' is not sample-specific; the adjacency matrix is fixed across all samples of a cancer type, and only the DEG set varies per patient. The phrase 'sample-independent random walks on patient-specific subnetworks' is therefore slightly misleading. Please clarify this point in the text.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the derivation is self-contained and the evaluation uses external gold standards without fitting to them.

full rationale

The paper's derivation chain does not reduce to its own inputs. Stage I scores are computed from somatic mutations, expression data, PPI/GRN networks, and optionally a WeSME mutual-exclusivity network; none of these inputs contains the CGC Tier 1 or IntOGen labels used for evaluation. The method is unsupervised and no parameter is fitted to the label set, so there is no fitted-input-called-prediction pattern. Stage II hypergraph refinement takes Stage I scores as node weights (Eqs. 19 and 25) and propagates them through hypergraph random walks; this is internal evidence propagation, not circularity against an external benchmark. The optional ME-network prior is inferred from the same TCGA mutation matrix, but it is used only as a restart prior and does not encode the evaluation labels. The concern that known drivers are network hubs and that Stage I random walks reward hubness is a potential confounding explanation for the reported performance, not a circularity: the output is not equal to an input by construction, and the gold standard is external. The Discussion's stated limitations (static PPI, pairwise GRN, no directionality, no CNV integration) concern biological validity and scope, not circular reasoning. No load-bearing self-citation, no imported uniqueness theorem, and no ansatz smuggled in via citation were found.

Assumptions & free parameters 11 free parameters · 8 assumptions · 0 invented entities

The paper's contribution is architectural: it combines known random-walk and hypergraph operators with 11 hand-set parameters and an ME prior inferred (via WeSME) from the same TCGA mutation data being analyzed. The biologically load-bearing inputs are three assumptions: only mutated genes can be drivers; drivers are expression-perturbing network hubs; and CGC Tier 1 is a valid gold standard. The mathematics (RWR convergence, hypergraph diffusion) is standard and correctly cited.

free parameters (11)
  • λ (PPI/GRN walk mixing) = 0.8
    Mixing weight between PPI and GRN transition matrices, §3.3.1 Eqs. 1-2; hand-set, no sensitivity analysis.
  • θ (Stage I restart probability) = 0.85
    Restart probability in RWR, §3.3.1-3.3.2 Eqs. 10 and 15; hand-set.
  • γ (score fusion weight) = 0.8
    Balance between static-network and co-expression scores after quantile normalization, Eq. 16; hand-set.
  • β (ME centrality balance) = 0.8
    Weight of betweenness vs harmonic centrality in the ME prior, Eq. 6; hand-set.
  • τ (co-expression correlation threshold) = 0.5
    Absolute-correlation cutoff for the dynamic co-expression adjacency matrix, Eq. 11; hand-set.
  • Top-K DEGs per sample = 500
    Cutoff for the DEG set used to induce the dynamic subnetwork, §3.3.2; hand-set.
  • σ (hyperedge bandwidth) = 0.1
    Bandwidth of the Gaussian similarity kernel for hyperedge weights, Eq. 20; hand-set; determines how many neighbor samples actually contribute.
  • α (Stage II restart probability) = 0.85
    Restart probability in hypergraph RWR, Eqs. 23 and 27; hand-set.
  • ME prior activation cutoff |V_m| = 200
    ME network used only when it has ≥200 genes, §3.3.1; hand-set; for 7 of 12 cancer types the prior is simply uniform.
  • N_c (personalized evaluation horizon) = 2 × median #CGC drivers
    Per-cancer-type top-K horizon for personal evaluation, Eq. 31; set from ground-truth driver counts — an evaluation parameter fitted to the label set.
  • Gene-length adjustment form = exp(-max(L_g-10^5,0)/10^4)
    Hand-chosen functional form down-weighting ultra-long genes, Eq. 8; constants (10^5, 10^4) are ad hoc.
assumptions (8)
  • domain assumption Only mutated genes are candidate drivers
    The ranked candidate set is each patient's mutated gene set MG (§3.3); drivers acting through CNV, methylation, or undetected mutations are invisible. Also gates the evaluation: genes outside the mutated set rank last (§3.5).
  • domain assumption Known driver genes are topological hubs
    §2.2 (Fig. S1a,b) shows CGC Tier 1 drivers have higher PPI degree and GRN out-degree; this is the stated rationale for the Stage I PPI/GRN walks. If undiscovered drivers are not hubs, the method will bias toward well-connected passengers.
  • domain assumption Driver mutations cause measurable transcriptional dysregulation in the same tumor sample
    Stage I dynamic walk (§3.3.2) connects mutated genes to the top-500 |z|-score DEGs; drivers that do not perturb expression generate no signal in the co-expression channel.
  • domain assumption CGC Tier 1 is a valid and sufficiently complete ground truth
    All headline metrics (§3.5, Table S2) use CGC Tier 1 as the reference. Incompleteness or bias toward well-studied hub genes directly shapes the reported precision and the IntOGen specificity analysis.
  • standard math Random walks with restart converge to a unique stationary distribution
    Eqs. 10, 15, 23, 27 assume classical RWR convergence; the transition matrices are stochastic and the restart terms are positive, so convergence is standard.
  • standard math Hypergraph random-walk formulation of Carletti et al. (2020) is applicable
    Stage II transitions (Eqs. 21-22, 26) adopt the two-step node→hyperedge→node formulation without modification.
  • domain assumption WeSME mutual-exclusivity calls at the chosen thresholds are reliable
    The ME prior (§3.3.1) trusts p≤0.01/FDR≤0.125 (H-H) and p≤0.001/FDR≤0.125 (R-H); false ME edges mis-seed the restart distribution, and the prior is used for only 5 of 12 cancer types (|V_m|≥200).
  • domain assumption Pearson correlation between expression profiles is a valid sample-similarity measure
    Hyperedge weights in personalized Stage II use exp(-(1-ρ_ij)²/2σ²) (Eq. 20); with σ=0.1 the kernel is sharply peaked, so non-near-identical neighbors contribute almost nothing.

how reviews work

0 comments
Cite this review

Pith. "Pith review of HyperNetWalk: A Unified Framework for Personalized and Cohort-Level Cancer Driver Gene Identification via Reverse Inference on Layered Signaling-Regulatory Network." pith.science (2026). https://pith.science/paper/WWC76AL7

@misc{pith2026260101337,
  author       = {Pith},
  title        = {Pith review of: HyperNetWalk: A Unified Framework for Personalized and Cohort-Level Cancer Driver Gene Identification via Reverse Inference on Layered Signaling-Regulatory Network},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/WWC76AL7}},
  note         = {Machine review of arXiv:2601.01337}
}
read the original abstract

Accurate identification of cancer driver genes from passenger mutations is essential for understanding tumorigenesis and clinical translation. We present HyperNetWalk, an unsupervised framework that unifies personalized and cohort-level driver gene identification within a shared inference architecture. HyperNetWalk builds a layered signaling-regulatory network by integrating protein-protein interactions, approximating upstream signaling, with a gene regulatory network for downstream transcriptional regulation, with transcription factors serving as interface nodes. Driver identification is formulated as an inverse problem in which observed transcriptional dysregulation is traced back to candidate upstream drivers by reverse random walk. The resulting sample-specific scores are used directly for personalized prediction and as node weights for cross-sample refinement through hypergraph random walk, enabling both local personalized and global cohort-level prediction. Across 12 TCGA cancer types, HyperNetWalk outperformed representative existing methods at both prediction levels. Ablation analyses supported the contributions of the reverse inference formulation and layered network architecture. Further analyses showed that HyperNetWalk captured cancer-type-specific driver signals, prioritized both recurrent and low-frequency candidate drivers, and produced predictions supported by drug-gene interaction and clinical actionability annotations.

Figures

Figures reproduced from arXiv: 2601.01337 by the authors.

Figure 1
Figure 1. Overview of HyperNetWalk. (a) HyperNetWalk integrates a somatic mutation matrix, gene expression matrix from tumor and normal samples, a protein–protein interaction (PPI) network, and a gene regulatory network (GRN) as inputs. (b) In Stage I, sample-independent random walks with restart are performed on patient-specific subnetworks derived from the PPI, GRN, and dynamic co-expression networks to generate initial per… view at source ↗
Figure 2
Figure 2. Personalized prediction performance: Precision@K across cancer types. [PITH_FULL_IMAGE:figures/full_fig_p006_2.png] view at source ↗
Figure 3
Figure 3. Cohort-level prediction performance: Precision@K across cancer types. [PITH_FULL_IMAGE:figures/full_fig_p007_3.png] view at source ↗
Figures from the paper (1 more)
Figure 4
Figure 4. Figure 4: Evaluation of cancer-type specificity using IntOGen driver genes. [PITH_FULL_IMAGE:figures/full_fig_p008_4.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

24 extracted references

  1. [1]

    Lawrence, Petar Stojanov, Paz Polak, Gregory V

    Michael S. Lawrence, Petar Stojanov, Paz Polak, Gregory V . Kryukov, Kristian Cibulskis, Andrey Sivachenko, Scott L. Carter, Chip Stewart, Craig H. Mermel, Steven A. Roberts, et al. Mutational heterogeneity in cancer and the search for new cancer-associated genes.Nature, 499(7457):214–218, 2013

    2013

  2. [2]

    Driverml: a machine learning algorithm for identifying driver genes in cancer sequencing studies

    Yi Han, Juze Y ang, Xinyi Qian, Wei-Chung Cheng, Shu-Hsuan Liu, Xing Hua, Liyuan Zhou, Y aning Y ang, Qingbiao Wu, Pengyuan Liu, et al. Driverml: a machine learning algorithm for identifying driver genes in cancer sequencing studies. Nucleic Acids Research, 47(8):e45–e45, 2019

    2019

  3. [3]

    Oncodrivefml: a general framework to identify coding and non-coding regions with cancer driver mutations

    Loris Mularoni, Radhakrishnan Sabarinathan, Jordi Deu-Pons, Abel Gonzalez-Perez, and Núria López-Bigas. Oncodrivefml: a general framework to identify coding and non-coding regions with cancer driver mutations. Genome biology, 17(1):128, 2016

    2016

  4. [4]

    Identification of cancer driver genes based on nucleotide context

    Felix Dietlein, Donate Weghorn, Amaro Taylor-Weiner, André Richters, Brendan Reardon, David Liu, Eric S Lander, Eliezer M Van Allen, and Shamil R Sunyaev. Identification of cancer driver genes based on nucleotide context. Nature genetics, 52(2):208–218, 2020

    2020

  5. [5]

    Integration of multi- omics data with graph convolutional networks to identify new cancer genes and their associated molecular mechanisms

    Roman Schulte-Sasse, Stefan Budach, Denes Hnisz, and Annalisa Marsico. Integration of multi- omics data with graph convolutional networks to identify new cancer genes and their associated molecular mechanisms. Nature Machine Intelligence, 3(6):513–526, 2021

    2021

  6. [6]

    Imi-driver: integrating multi-level gene networks and multi-omics for cancer driver gene identification

    Peiting Shi, Junmin Han, Yinghao Zhang, Guanpu Li, and Xionghui Zhou. Imi-driver: integrating multi-level gene networks and multi-omics for cancer driver gene identification. PLOS Computa- tional Biology, 20(8):e1012389, 2024

    2024

  7. [7]

    Pan-cancer network analysis identifies combinations of rare somatic mutations across pathways and protein complexes

    Mark DM Leiserson, Fabio Vandin, Hsin-Ta Wu, Jason R Dobson, Jonathan V Eldridge, Jacob L Thomas, Alexandra Papoutsaki, Y ounhun Kim, Beifang Niu, Michael McLellan, et al. Pan-cancer network analysis identifies combinations of rare somatic mutations across pathways and protein complexes. Nature genetics, 47(2):106–114, 2015

    2015

  8. [8]

    Drivernet: uncovering the impact of somatic driver mutations on transcriptional networks in cancer

    Ali Bashashati, Gholamreza Haffari, Jiarui Ding, Gavin Ha, Kenneth Lui, Jamie Rosner, David G Huntsman, Carlos Caldas, Samuel A Aparicio, and Sohrab P Shah. Drivernet: uncovering the impact of somatic driver mutations on transcriptional networks in cancer. Genome biology, 13(12): R124, 2012

    2012

Show all 24 references
  1. [9]

    Velculescu, Shibin Zhou, Luis A

    Bert Vogelstein, Nickolas Papadopoulos, Victor E. Velculescu, Shibin Zhou, Luis A. Diaz, and Kenneth W. Kinzler. Cancer genome landscapes. Science, 339(6127):1546–1558, 2013

    2013

  2. [10]

    Dawnrank: discovering personalized driver genes in cancer

    Jack P Hou and Jian Ma. Dawnrank: discovering personalized driver genes in cancer. Genome medicine, 6(7):56, 2014. HyperNetWalk: A Unified Framework 20

    2014

  3. [11]

    Discovering personalized driver mutation profiles of single samples in cancer by network control strategy

    Wei-Feng Guo, Shao-Wu Zhang, Li-Li Liu, Fei Liu, Qian-Qian Shi, Lei Zhang, Ying Tang, Tao Zeng, and Luonan Chen. Discovering personalized driver mutation profiles of single samples in cancer by network control strategy. Bioinformatics, 34(11):1893–1903, 2018

    1903

  4. [12]

    Personadrive: a method for the identification and prioritization of personalized cancer drivers

    Cesim Erten, Aissa Houdjedj, Hilal Kazan, and Ahmed Amine Taleb Bahmed. Personadrive: a method for the identification and prioritization of personalized cancer drivers. Bioinformatics, 38 (13):3407–3414, 2022

    2022

  5. [13]

    A novel hypergraph model for identifying and prioritizing personalized drivers in cancer

    Naiqian Zhang, Fubin Ma, Dong Guo, Yuxuan Pang, Chenye Wang, Yusen Zhang, Xiaoqi Zheng, and Mingyi Wang. A novel hypergraph model for identifying and prioritizing personalized drivers in cancer. PLOS Computational Biology, 20(4):e1012068, 2024

    2024

  6. [14]

    Drivermp enables improved identification of cancer driver genes

    Y angyang Liu, Jiyun Han, Tongxin Kong, Nannan Xiao, Qinglin Mei, and Juntao Liu. Drivermp enables improved identification of cancer driver genes. GigaScience, 12:giad106, 2023

    2023

  7. [15]

    Pitch: A pathway-induced prioritization of personalized cancer driver genes based on higher-order interactions

    Yuhe Wang, Suoqin Jin, and Xiufen Zou. Pitch: A pathway-induced prioritization of personalized cancer driver genes based on higher-order interactions. IEEE Journal of Biomedical and Health Informatics, 2025

    2025

  8. [16]

    Prodigy: personalized prioritization of driver genes

    Gal Dinstag and Ron Shamir. Prodigy: personalized prioritization of driver genes. Bioinformatics, 36(6):1831–1839, 2020

    2020

  9. [17]

    Driver- rwh: discovering cancer driver genes by random walk on a gene mutation hypergraph

    Chenye Wang, Junhan Shi, Jiansheng Cai, Yusen Zhang, Xiaoqi Zheng, and Naiqian Zhang. Driver- rwh: discovering cancer driver genes by random walk on a gene mutation hypergraph. BMC bioin- formatics, 23(1):277, 2022

    2022

  10. [18]

    A random walk-based method to identify driver genes by integrating the subcellular localization and variation frequency into bipartite graph

    Junrong Song, Wei Peng, and Feng Wang. A random walk-based method to identify driver genes by integrating the subcellular localization and variation frequency into bipartite graph. BMC bioin- formatics, 20(1):238, 2019

    2019

  11. [19]

    Visualizing and interpreting cancer genomics data via the xena platform

    Mary J Goldman, Brian Craft, Mim Hastie, Kristupas Repečka, Fran McDade, Akhil Kamath, Ayan Banerjee, Yunhai Luo, Dave Rogers, Angela N Brooks, et al. Visualizing and interpreting cancer genomics data via the xena platform. Nature biotechnology, 38(6):675–678, 2020

    2020

  12. [20]

    The string database in 2023: protein–protein association networks and functional enrichment analyses for any sequenced genome of interest

    Damian Szklarczyk, Rebecca Kirsch, Mikaela Koutrouli, Katerina Nastou, Farrokh Mehryary, Radja Hachilif, Annika L Gable, Tao Fang, Nadezhda T Doncheva, Sampo Pyysalo, et al. The string database in 2023: protein–protein association networks and functional enrichment analyses fo...

    2023

  13. [21]

    Regnetwork 2025: an in- tegrative data repository for gene regulatory networks in human and mouse

    Bowen Li, Chuanyuan Wang, Yijuan Wang, Pengpai Li, and Zhi-Ping Liu. Regnetwork 2025: an in- tegrative data repository for gene regulatory networks in human and mouse. Nucleic Acids Research, page gkaf779, 2025

    2025

  14. [22]

    Wesme: uncovering mutual exclusivity of cancer drivers and beyond

    Y oo-Ah Kim, Sanna Madan, and Teresa M Przytycka. Wesme: uncovering mutual exclusivity of cancer drivers and beyond. Bioinformatics, 33(6):814–821, 2017

    2017

  15. [23]

    Random walks on hyper- graphs

    Timoteo Carletti, Federico Battiston, Giulia Cencetti, and Duccio Fanelli. Random walks on hyper- graphs. Physical review E, 101(2):022308, 2020

    2020

  16. [24]

    The cosmic cancer gene census: describing genetic dysfunction across all human cancers

    Zbyslaw Sondka, Sally Bamford, Charlotte G Cole, Sari A Ward, Ian Dunham, and Simon A Forbes. The cosmic cancer gene census: describing genetic dysfunction across all human cancers. Nature Reviews Cancer, 18(11):696–705, 2018. HyperNetWalk: A Unified Framework 21 Supplementary...

    2018

Pith tools

Reviewed August 3, 2026 · model on record in the stance chip above.