Local graph estimation with pathwise false discovery control

David B. Dunson; Jeffrey W. Miller; Noureddine Melikechi; Omar Melikechi

arxiv: 2507.17172 · v2 · pith:K5GPS4R7new · submitted 2025-07-23 · 📊 stat.ME · stat.AP

Local graph estimation with pathwise false discovery control

Omar Melikechi , David B. Dunson , Noureddine Melikechi , Jeffrey W. Miller This is my paper

Pith reviewed 2026-05-19 03:30 UTC · model grok-4.3

classification 📊 stat.ME stat.AP

keywords local graph estimationpathwise feature selectionfalse discovery controlnetwork inferencevariable selectionmixed data typesnonlinear dependencies

0 comments

The pith

Pathwise feature selection recovers local subgraphs around target variables with finite-sample false discovery control.

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

Many datasets center on a few key variables whose connections to the rest of the system matter most, yet estimating the entire network often buries the relevant local patterns. The paper introduces local graph estimation as a targeted alternative and shows that standard full-network methods frequently miss these structures. It presents pathwise feature selection, which works by repeatedly applying feature selection and carrying uncertainty forward along possible paths. This yields valid false-discovery guarantees that hold in finite samples even when variables are of mixed types or linked by nonlinear relations. Real-data examples in environmental health, multiomics, brain imaging, and RNA sequencing produce networks that match known biology and suggest new questions.

Core claim

The central claim is that pathwise feature selection estimates local subgraphs by iteratively applying feature selection and propagating uncertainty along network paths, providing rigorous finite-sample false discovery control even in settings with mixed variable types and nonlinear dependencies.

What carries the argument

Pathwise feature selection (PFS), an iterative procedure that selects features and propagates uncertainty along network paths to control local false discoveries.

If this is right

Local structures around scientifically interesting targets can be recovered without first estimating the full network.
The procedure remains valid for mixed continuous, discrete, and categorical variables linked by nonlinear relationships.
Recovered local networks align with established domain mechanisms in environmental, multiomics, and neuroscience settings.
The same framework can generate testable new hypotheses by highlighting previously unrecognized connections.

Where Pith is reading between the lines

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

The approach could be extended to time-varying or longitudinal targets by propagating uncertainty across time steps.
If the pathwise propagation step can be made computationally lighter, the method becomes practical for very large variable sets.
Comparison with existing local-inference techniques would clarify whether the finite-sample guarantee is the main practical advantage.

Load-bearing premise

Iterative feature selection plus pathwise uncertainty propagation produces valid finite-sample false discovery control without further conditions on the dependence structure or variable types.

What would settle it

A simulation study in which the true local graph is known and the method reports a positive discovery rate that exceeds the nominal control level after accounting for the declared error rate.

Figures

Figures reproduced from arXiv: 2507.17172 by David B. Dunson, Jeffrey W. Miller, Noureddine Melikechi, Omar Melikechi.

**Figure 1.** Figure 1: Benefits of pathwise false discovery control and limitations of existing methods on simulated and real data. (a) The true local graph of radius 3 around the target variable X1, shown in yellow. (b) Pathwise feature selection identifies 21 true edges and 3 false edges. It also provides edge-specific uncertainty quantification in the form of q-values—shown as edge weights—with lower q-values indicating great… view at source ↗

**Figure 2.** Figure 2: Radius 2 graph around county-level cancer incidence and mortality, estimated by PFS. The target variables, cancer incidence and mortality, are shown in yellow. Nodes directly connected to a target are shown in green, and nodes connected to a green node but not to either target are shown in purple. Edges are annotated with their q-values, with smaller values indicating stronger evidence of conditional depen… view at source ↗

**Figure 3.** Figure 3: Geographic distributions of cancer burden and selected environmental and socioeconomic variables across the contiguous United States. Heatmaps show county-level values for cancer incidence and mortality (top row) and selected environmental exposures (bottom left) and social and demographic factors (bottom right). Cancer incidence tends to align with elevated levels of exposures—including particulate matte… view at source ↗

**Figure 4.** Figure 4: Radius 3 graph around clinical variables in TCGA breast cancer data, estimated by PFS. The target variables—histological type, pathologic stage, and survival status—are shown in yellow. Nodes directly connected to a target are shown in green, nodes at distance two are shown in blue, and nodes at distance three are shown in purple. Edges are annotated with q-values, with smaller values indicating stronger e… view at source ↗

read the original abstract

Many datasets include a small set of variables, such as biomarkers or clinical outcomes, whose relationships to the broader system are of primary scientific interest. Estimating the full network of inter-variable relationships in such settings often obscures local structures around these targets, limiting interpretability. To address this fundamental problem, we introduce local graph estimation, a statistical framework for inferring substructures around target variables. We show that traditional graph estimation methods often fail to recover local structure, and present pathwise feature selection (PFS) as an effective alternative. PFS estimates local subgraphs by iteratively applying feature selection and propagating uncertainty along network paths, providing rigorous finite-sample false discovery control even in settings with mixed variable types and nonlinear dependencies. In four distinct applications spanning environmental and public health, multiomics, brain connectomics, and single-nucleus RNA sequencing, PFS recovers interpretable networks consistent with domain knowledge, highlighting its ability to uncover established mechanisms and generate novel hypotheses.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Desk Editor's Note private letter to a colleague

PFS provides a focused alternative for local network estimation with finite-sample FDR control, though the general guarantees deserve close verification.

read the letter

The thing to know is that this paper gives a way to estimate just the local connections around a few target variables while claiming rigorous finite-sample control on false discoveries. It does this through pathwise feature selection, which applies selection iteratively and carries uncertainty forward along paths. That is new compared to standard full-graph methods that try to recover everything at once. The applications are a strength. The authors test it on environmental data, multiomics, brain imaging, and single-cell RNA, and the recovered networks match known biology while pointing to new hypotheses. Seeing it work across domains helps. The soft spot is the finite-sample FDR claim in the general setting. The stress-test note is right to ask whether the propagation step preserves the bound without extra assumptions on linearity or residual independence. If the paper only proves it under more restrictive conditions than advertised, the guarantee does not fully extend to the nonlinear mixed-type case. I would check the theorem statement and any supporting simulations carefully. This paper is for people who need focused network estimates rather than complete graphs, such as biologists or environmental scientists with a few key outcomes. It should interest readers working on selective inference or graphical models with error control. The idea is coherent and the empirical results are solid enough that it deserves a serious referee. I would send it to peer review with a note to the referees to pay special attention to the conditions required for the finite-sample control to hold in the claimed generality.

Referee Report

2 major / 2 minor

Summary. The paper introduces local graph estimation as a framework for inferring substructures around target variables of primary interest, rather than full networks. It proposes pathwise feature selection (PFS), which iteratively applies feature selection and propagates uncertainty along network paths. The central claim is that PFS delivers rigorous finite-sample false discovery control for local subgraphs, even with mixed variable types and nonlinear dependencies. The method is demonstrated in four applications (environmental/public health, multiomics, brain connectomics, single-nucleus RNA sequencing) where it recovers interpretable networks consistent with domain knowledge.

Significance. If the finite-sample FDR control holds under the advertised conditions, the work would offer a valuable tool for improving interpretability in high-dimensional network data by focusing statistical effort on local structures around key targets. The applications across diverse domains provide evidence of practical utility, though the theoretical advance would be the primary contribution.

major comments (2)

[Abstract and §4] Abstract and §4 (Theoretical Results): The manuscript asserts 'rigorous finite-sample false discovery control' via pathwise propagation but provides no theorem statement, proof sketch, or derivation of the bound. Without explicit conditions on the iterative selection procedure, test statistics, or dependence structure, it is impossible to verify whether the control holds in the general nonlinear/mixed-type regime claimed.
[§3.2] §3.2 (Pathwise Uncertainty Propagation): The description of propagating uncertainty along paths does not clarify how the overall FDR bound is preserved when combining multiple iterative selections; if the mechanism relies on union bounds or recursive conditioning that implicitly require linearity, residual independence, or specific variable types, the finite-sample guarantee fails to extend to the nonlinear and mixed-type settings advertised as a strength.

minor comments (2)

[Introduction] The abstract and introduction could more explicitly contrast PFS with existing local network methods to better position the novelty.
[Applications] Figure captions in the applications section would benefit from additional detail on how domain knowledge is used to interpret recovered edges.

Simulated Author's Rebuttal

2 responses · 0 unresolved

We thank the referee for the careful reading and constructive comments. We address each major point below and will revise the manuscript to strengthen the presentation of the theoretical results.

read point-by-point responses

Referee: [Abstract and §4] Abstract and §4 (Theoretical Results): The manuscript asserts 'rigorous finite-sample false discovery control' via pathwise propagation but provides no theorem statement, proof sketch, or derivation of the bound. Without explicit conditions on the iterative selection procedure, test statistics, or dependence structure, it is impossible to verify whether the control holds in the general nonlinear/mixed-type regime claimed.

Authors: We agree that an explicit theorem statement and proof sketch are needed for full verification. In the revised manuscript we will add a formal theorem in Section 4 that states the finite-sample FDR control for PFS under the stated conditions on the base feature selection procedure. The proof proceeds by induction on path length: each iterative selection step inherits marginal validity from the base method (which is valid for mixed types and nonlinear dependence via appropriate scores or tests), and the pathwise adjustment applies a conservative threshold derived from a union bound over the number of paths considered. This construction does not require linearity or residual independence and therefore extends to the general regime claimed. A detailed proof sketch will be included. revision: yes
Referee: [§3.2] §3.2 (Pathwise Uncertainty Propagation): The description of propagating uncertainty along paths does not clarify how the overall FDR bound is preserved when combining multiple iterative selections; if the mechanism relies on union bounds or recursive conditioning that implicitly require linearity, residual independence, or specific variable types, the finite-sample guarantee fails to extend to the nonlinear and mixed-type settings advertised as a strength.

Authors: The propagation mechanism preserves the bound through recursive conditioning on prior selections combined with a path-length-adjusted threshold that remains valid under arbitrary dependence. At each step the effective significance level is tightened by a factor that accounts for the maximum number of paths reaching that node; because the adjustment is distribution-free and only uses the marginal validity of the base selection procedure, it does not invoke linearity or independence assumptions. We will expand §3.2 with an explicit description of this recursive argument and a small numerical illustration on nonlinear data to demonstrate that the overall FDR remains controlled. revision: yes

Circularity Check

0 steps flagged

No circularity: PFS derives finite-sample FDR control via independent propagation steps

full rationale

The paper presents pathwise feature selection (PFS) as a novel iterative procedure that applies feature selection and propagates uncertainty along paths to achieve finite-sample false discovery control for local subgraphs. The abstract and description frame this as a statistical framework with rigorous guarantees even under mixed types and nonlinear dependencies, without any quoted equations, fitted parameters renamed as predictions, or self-citations that reduce the control result to its own inputs by construction. The derivation chain is self-contained because the claimed control follows from the propagation mechanism applied to standard selection steps, rather than tautologically re-expressing a fitted quantity or prior result from the same authors.

Axiom & Free-Parameter Ledger

0 free parameters · 0 axioms · 0 invented entities

Abstract-only review provides no explicit free parameters, axioms, or invented entities; typical feature selection methods implicitly rely on regularization parameters and independence assumptions that are not detailed here.

pith-pipeline@v0.9.0 · 5697 in / 1021 out tokens · 32197 ms · 2026-05-19T03:30:06.484108+00:00 · methodology

discussion (0)

Reference graph

Works this paper leans on

64 extracted references · 64 canonical work pages

[1]

author Eicher, T. et al. title Metabolomics and multi-omics integration: a survey of computational methods and resources . journal Metabolites volume 10 , pages 202 ( year 2020 )

work page 2020
[2]

author Maitre, L. et al. title State-of-the-art methods for exposure-health studies: R esults from the exposome data challenge event . journal Environment International volume 168 , pages 107422 ( year 2022 )

work page 2022
[3]

author Allen, G. I. , author Gan, L. & author Zheng, L. title Interpretable machine learning for discovery: Statistical challenges and opportunities . journal Annual Review of Statistics and its Application volume 11 ( year 2023 )

work page 2023
[4]

author Chen, V. et al. title Applying interpretable machine learning in computational biology-pitfalls, recommendations and opportunities for new developments . journal Nature Methods volume 21 , pages 1454--1461 ( year 2024 )

work page 2024
[5]

& author Mukhopadhyay, A

author Acharya, D. & author Mukhopadhyay, A. title A comprehensive review of machine learning techniques for multi-omics data integration: challenges and applications in precision oncology . journal Briefings in Functional Genomics volume 23 , pages 549--560 ( year 2024 )

work page 2024
[6]

author Warkiani, M. E. & author Moattar, M. H. title A comprehensive survey on recent feature selection methods for mixed data: Challenges, solutions and future directions . journal Neurocomputing pages 129372 ( year 2025 )

work page 2025
[7]

author Mallick, H. et al. title Multivariable association discovery in population-scale meta-omics studies . journal PLoS Computational Biology volume 17 , pages e1009442 ( year 2021 )

work page 2021
[8]

author Rahnenf \"u hrer, J. et al. title Statistical analysis of high-dimensional biomedical data: a gentle introduction to analytical goals, common approaches and challenges . journal BMC Medicine volume 21 , pages 182 ( year 2023 )

work page 2023
[9]

, author Wohlfart, C

author Duroux, D. , author Wohlfart, C. , author Van Steen, K. , author Vladimirova, A. & author King, M. title Graph-based multi-modality integration for prediction of cancer subtype and severity . journal Scientific Reports volume 13 , pages 19653 ( year 2023 )

work page 2023
[10]

author Young, A. S. et al. title The need for a cancer exposome atlas: a scoping review . journal JNCI Cancer Spectrum volume 9 , pages pkae122 ( year 2025 )

work page 2025
[11]

author Cantini, L. et al. title Benchmarking joint multi-omics dimensionality reduction approaches for the study of cancer . journal Nature communications volume 12 , pages 124 ( year 2021 )

work page 2021
[12]

, author Das, V

author Krassowski, M. , author Das, V. , author Sahu, S. K. & author Misra, B. B. title State of the field in multi-omics research: from computational needs to data mining and sharing . journal Frontiers in Genetics volume 11 , pages 610798 ( year 2020 )

work page 2020
[13]

author Boehm, K. M. , author Khosravi, P. , author Vanguri, R. , author Gao, J. & author Shah, S. P. title Harnessing multimodal data integration to advance precision oncology . journal Nature Reviews Cancer volume 22 , pages 114--126 ( year 2022 )

work page 2022
[14]

author Rudin, C. et al. title Interpretable machine learning: Fundamental principles and 10 grand challenges . journal Statistics Surveys volume 16 , pages 1--85 ( year 2022 )

work page 2022
[15]

& author Armon, A

author Shwartz-Ziv, R. & author Armon, A. title Tabular data: Deep learning is not all you need . journal Information Fusion volume 81 , pages 84--90 ( year 2022 )

work page 2022
[16]

, author Oyallon, E

author Grinsztajn, L. , author Oyallon, E. & author Varoquaux, G. title Why do tree-based models still outperform deep learning on typical tabular data? journal Advances in Neural Information Processing Systems volume 35 , pages 507--520 ( year 2022 )

work page 2022
[17]

& author Maathuis, M

author Drton, M. & author Maathuis, M. H. title Structure learning in graphical modeling . journal Annual Review of Statistics and its Application volume 4 , pages 365--393 ( year 2017 )

work page 2017
[18]

& author Friedman, N

author Koller, D. & author Friedman, N. title Probabilistic Graphical Models: Principles and Techniques ( publisher MIT press , year 2009 )

work page 2009
[19]

author Weinstein, J. N. et al. title The cancer genome atlas pan-cancer analysis project . journal Nature Genetics volume 45 , pages 1113--1120 ( year 2013 )

work page 2013
[20]

, author Hastie, T

author Friedman, J. , author Hastie, T. & author Tibshirani, R. title Sparse inverse covariance estimation with the graphical lasso . journal Biostatistics volume 9 , pages 432--441 ( year 2008 )

work page 2008
[21]

& author B \"u hlmann, P

author Meinshausen, N. & author B \"u hlmann, P. title High-dimensional graphs and variable selection with the lasso . journal The Annals of Statistics volume 34 , pages 1436--1462 ( year 2006 )

work page 2006
[22]

& author Miller, J

author Melikechi, O. & author Miller, J. W. title Integrated path stability selection . journal Journal of the American Statistical Association ( year 2025 )

work page 2025
[23]

, author Dunson, D

author Melikechi, O. , author Dunson, D. B. & author Miller, J. W. title Nonparametric IPSS: fast, flexible feature selection with false discovery control . journal Bioinformatics volume 41 , pages btaf299 ( year 2025 )

work page 2025
[24]

, author Sechidis, K

author Nogueira, S. , author Sechidis, K. & author Brown, G. title On the stability of feature selection algorithms . journal Journal of Machine Learning Research volume 18 , pages 1--54 ( year 2018 )

work page 2018
[25]

author Maitre, L. et al. title Multi-omics signatures of the human early life exposome . journal Nature communications volume 13 , pages 7024 ( year 2022 )

work page 2022
[26]

title Metabolomics and the multi-omics view of cancer

author Wishart, D. title Metabolomics and the multi-omics view of cancer . journal Metabolites volume 12 , pages 154 ( year 2022 )

work page 2022
[27]

author Messer, L. C. , author Jagai, J. S. , author Rappazzo, K. M. & author Lobdell, D. T. title Construction of an environmental quality index for public health research . journal Environmental Health volume 13 , pages 1--22 ( year 2014 )

work page 2014
[28]

, author Egerter, S

author Braveman, P. , author Egerter, S. & author Williams, D. R. title The social determinants of health: coming of age . journal Annual Review of Public Health volume 32 , pages 381--398 ( year 2011 )

work page 2011
[29]

, author Davis, J

author Wei, Y. , author Davis, J. & author Bina, W. F. title Ambient air pollution is associated with the increased incidence of breast cancer in us . journal International Journal of Environmental Health Research volume 22 , pages 12--21 ( year 2012 )

work page 2012
[30]

author Wang, N. et al. title Lung cancer and particulate pollution: A critical review of spatial and temporal analysis evidence . journal Environmental Research volume 164 , pages 585--596 ( year 2018 )

work page 2018
[31]

author Lipfert, F. W. & author Wyzga, R. E. title Longitudinal relationships between lung cancer mortality rates, smoking, and ambient air quality: a comprehensive review and analysis . journal Critical Reviews in Toxicology volume 49 , pages 790--818 ( year 2019 )

work page 2019
[32]

author Kentros, P. A. et al. title Ambient particulate matter air pollution exposure and ovarian cancer incidence in the usa: an ecological study . journal BJOG: An International Journal of Obstetrics & Gynaecology volume 131 , pages 690--698 ( year 2024 )

work page 2024
[33]

author Organization, W. H. title WHO global air quality guidelines: Particulate matter (PM _ 2.5 and PM _ 10 ), ozone, nitrogen dioxide, sulfur dioxide and carbon monoxide ( publisher World Health Organization , year 2021 )

work page 2021
[34]

, author Zhao, C

author Luo, Y. , author Zhao, C. & author Chen, F. title Multiomics research: Principles and challenges in integrated analysis . journal BioDesign Research volume 6 , pages 0059 ( year 2024 )

work page 2024
[35]

, author Seldin, M

author Hasin, Y. , author Seldin, M. & author Lusis, A. title Multi-omics approaches to disease . journal Genome Biology volume 18 , pages 1--15 ( year 2017 )

work page 2017
[36]

author Hayes, C. N. , author Nakahara, H. , author Ono, A. , author Tsuge, M. & author Oka, S. title From omics to multi-omics: A review of advantages and tradeoffs . journal Genes volume 15 , pages 1551 ( year 2024 )

work page 2024
[37]

author Wekesa, J. S. & author Kimwele, M. title A review of multi-omics data integration through deep learning approaches for disease diagnosis, prognosis, and treatment . journal Frontiers in Genetics volume 14 , pages 1199087 ( year 2023 )

work page 2023
[38]

author Ganini, C. et al. title Global mapping of cancers: The Cancer Genome Atlas and beyond . journal Molecular Oncology volume 15 , pages 2823--2840 ( year 2021 )

work page 2021
[39]

author Mallick, H. et al. title An integrated bayesian framework for multi-omics prediction and classification . journal Statistics in Medicine volume 43 , pages 983--1002 ( year 2024 )

work page 2024
[40]

author Mohr, A. E. , author Ortega-Santos, C. P. , author Whisner, C. M. , author Klein-Seetharaman, J. & author Jasbi, P. title Navigating challenges and opportunities in multi-omics integration for personalized healthcare . journal Biomedicines volume 12 , pages 1496 ( year 2024 )

work page 2024
[41]

author Christgen, M. et al. title Lobular breast cancer: Clinical, molecular and morphological characteristics . journal Pathology-Research and Practice volume 212 , pages 583--597 ( year 2016 )

work page 2016
[42]

title The prognosis of invasive ductal carcinoma, lobular carcinoma and mixed ductal and lobular carcinoma according to molecular subtypes of the breast

author Zhao, H. title The prognosis of invasive ductal carcinoma, lobular carcinoma and mixed ductal and lobular carcinoma according to molecular subtypes of the breast . journal Breast Cancer volume 28 , pages 187--195 ( year 2021 )

work page 2021
[43]

& author Metzger-Filho, O

author Barroso-Sousa, R. & author Metzger-Filho, O. title Differences between invasive lobular and invasive ductal carcinoma of the breast: results and therapeutic implications . journal Therapeutic Advances in Medical Oncology volume 8 , pages 261--266 ( year 2016 )

work page 2016
[44]

, author Luo, J.-H

author Zhang, X.-X. , author Luo, J.-H. & author Wu, L.-Q. title FN1 overexpression is correlated with unfavorable prognosis and immune infiltrates in breast cancer . journal Frontiers in Genetics volume 13 , pages 913659 ( year 2022 )

work page 2022
[45]

author Noman, M. Z. et al. title Tumor-promoting effects of myeloid-derived suppressor cells are potentiated by hypoxia-induced expression of miR-210 . journal Cancer Research volume 75 , pages 3771--3787 ( year 2015 )

work page 2015
[46]

author McCormick, R. et al. title miR-210 is a target of hypoxia-inducible factors 1 and 2 in renal cancer, regulates ISCU and correlates with good prognosis . journal British Journal of Cancer volume 108 , pages 1133--1142 ( year 2013 )

work page 2013
[47]

author Yang, F. et al. title MiR-210 in exosomes derived from CAFs promotes non-small cell lung cancer migration and invasion through PTEN/PI3K/AKT pathway . journal Cellular Signalling volume 73 , pages 109675 ( year 2020 )

work page 2020
[48]

author Kojima, S. et al. title Tumour suppressors miR-1 and miR-133a target the oncogenic function of purine nucleoside phosphorylase (PNP) in prostate cancer . journal British Journal of Cancer volume 106 , pages 405--413 ( year 2012 )

work page 2012
[49]

& author Lilly, B

author Sawant, D. & author Lilly, B. title MicroRNA-145 targets in cancer and the cardiovascular system: evidence for common signaling pathways . journal Vascular Biology volume 2 , pages R115--R128 ( year 2020 )

work page 2020
[50]

author Sharma, N. et al. title BCR/ABL1 and BCR are under the transcriptional control of the MYC oncogene . journal Molecular Cancer volume 14 , pages 1--11 ( year 2015 )

work page 2015
[51]

, author Stern, A

author Yang, H.-C. , author Stern, A. & author Chiu, D. T.-Y. title G6PD: A hub for metabolic reprogramming and redox signaling in cancer . journal Biomedical Journal volume 44 , pages 285--292 ( year 2021 )

work page 2021
[52]

& author Hastie, T

author Mazumder, R. & author Hastie, T. title Exact covariance thresholding into connected components for large-scale graphical lasso . journal The Journal of Machine Learning Research volume 13 , pages 781--794 ( year 2012 )

work page 2012
[53]

, author Lafferty, J

author Liu, H. , author Lafferty, J. & author Wasserman, L. title The nonparanormal: S emiparametric estimation of high dimensional undirected graphs. journal Journal of Machine Learning Research volume 10 ( year 2009 )

work page 2009
[54]

& author Zou, H

author Xue, L. & author Zou, H. title Regularized rank-based estimation of high-dimensional nonparanormal graphical models . journal The Annals of Statistics volume 40 , pages 2541--2571 ( year 2012 )

work page 2012
[55]

title Gaussian graphical model estimation with false discovery rate control

author Liu, W. title Gaussian graphical model estimation with false discovery rate control . journal The Annals of Statistics volume 41 , pages 2948--2978 ( year 2013 )

work page 2013
[56]

, author Ren, Z

author Zhang, R. , author Ren, Z. & author Chen, W. title Silggm: An extensive r package for efficient statistical inference in large-scale gene networks . journal PLoS Computational Biology volume 14 , pages e1006369 ( year 2018 )

work page 2018
[57]

& author Maathuis, M

author Li, J. & author Maathuis, M. H. title GGM knockoff filter: False discovery rate control for Gaussian graphical models . journal Journal of the Royal Statistical Society Series B: Statistical Methodology volume 83 , pages 534--558 ( year 2021 )

work page 2021
[58]

author Storey, J. D. title The positive false discovery rate: A B ayesian interpretation and the q-value . journal The Annals of Statistics volume 31 , pages 2013--2035 ( year 2003 )

work page 2013
[59]

author Vasaikar, S. V. , author Straub, P. , author Wang, J. & author Zhang, B. title Linkedomics: analyzing multi-omics data within and across 32 cancer types . journal Nucleic Acids Research volume 46 , pages D956--D963 ( year 2018 )

work page 2018
[60]

& author Guestrin, C

author Chen, T. & author Guestrin, C. title Xgboost: A scalable tree boosting system . In booktitle Proceedings of the 22nd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining , pages 785--794 ( year 2016 )

work page 2016
[61]

, author Liu, H

author Zhao, T. , author Liu, H. , author Roeder, K. , author Lafferty, J. & author Wasserman, L. title The huge package for high-dimensional undirected graph estimation in r . journal The Journal of Machine Learning Research volume 13 , pages 1059--1062 ( year 2012 )

work page 2012
[62]

, author Sun, T

author Ren, Z. , author Sun, T. , author Zhang, C.-H. & author Zhou, H. H. title Asymptotic normality and optimalities in estimation of large Gaussian graphical models . journal The Annals of Statistics volume 43 , pages 991--1026 ( year 2015 )

work page 2015
[63]

& author Van De Geer, S

author Jankova, J. & author Van De Geer, S. title Confidence intervals for high-dimensional inverse covariance estimation . journal Electronic Journal of Statistics volume 9 , pages 1205--1229 ( year 2015 )

work page 2015
[64]

& author van de Geer, S

author Jankov \'a , J. & author van de Geer, S. title Honest confidence regions and optimality in high-dimensional precision matrix estimation . journal Test volume 26 , pages 143--162 ( year 2017 )

work page 2017

[1] [1]

author Eicher, T. et al. title Metabolomics and multi-omics integration: a survey of computational methods and resources . journal Metabolites volume 10 , pages 202 ( year 2020 )

work page 2020

[2] [2]

author Maitre, L. et al. title State-of-the-art methods for exposure-health studies: R esults from the exposome data challenge event . journal Environment International volume 168 , pages 107422 ( year 2022 )

work page 2022

[3] [3]

author Allen, G. I. , author Gan, L. & author Zheng, L. title Interpretable machine learning for discovery: Statistical challenges and opportunities . journal Annual Review of Statistics and its Application volume 11 ( year 2023 )

work page 2023

[4] [4]

author Chen, V. et al. title Applying interpretable machine learning in computational biology-pitfalls, recommendations and opportunities for new developments . journal Nature Methods volume 21 , pages 1454--1461 ( year 2024 )

work page 2024

[5] [5]

& author Mukhopadhyay, A

author Acharya, D. & author Mukhopadhyay, A. title A comprehensive review of machine learning techniques for multi-omics data integration: challenges and applications in precision oncology . journal Briefings in Functional Genomics volume 23 , pages 549--560 ( year 2024 )

work page 2024

[6] [6]

author Warkiani, M. E. & author Moattar, M. H. title A comprehensive survey on recent feature selection methods for mixed data: Challenges, solutions and future directions . journal Neurocomputing pages 129372 ( year 2025 )

work page 2025

[7] [7]

author Mallick, H. et al. title Multivariable association discovery in population-scale meta-omics studies . journal PLoS Computational Biology volume 17 , pages e1009442 ( year 2021 )

work page 2021

[8] [8]

author Rahnenf \"u hrer, J. et al. title Statistical analysis of high-dimensional biomedical data: a gentle introduction to analytical goals, common approaches and challenges . journal BMC Medicine volume 21 , pages 182 ( year 2023 )

work page 2023

[9] [9]

, author Wohlfart, C

author Duroux, D. , author Wohlfart, C. , author Van Steen, K. , author Vladimirova, A. & author King, M. title Graph-based multi-modality integration for prediction of cancer subtype and severity . journal Scientific Reports volume 13 , pages 19653 ( year 2023 )

work page 2023

[10] [10]

author Young, A. S. et al. title The need for a cancer exposome atlas: a scoping review . journal JNCI Cancer Spectrum volume 9 , pages pkae122 ( year 2025 )

work page 2025

[11] [11]

author Cantini, L. et al. title Benchmarking joint multi-omics dimensionality reduction approaches for the study of cancer . journal Nature communications volume 12 , pages 124 ( year 2021 )

work page 2021

[12] [12]

, author Das, V

author Krassowski, M. , author Das, V. , author Sahu, S. K. & author Misra, B. B. title State of the field in multi-omics research: from computational needs to data mining and sharing . journal Frontiers in Genetics volume 11 , pages 610798 ( year 2020 )

work page 2020

[13] [13]

author Boehm, K. M. , author Khosravi, P. , author Vanguri, R. , author Gao, J. & author Shah, S. P. title Harnessing multimodal data integration to advance precision oncology . journal Nature Reviews Cancer volume 22 , pages 114--126 ( year 2022 )

work page 2022

[14] [14]

author Rudin, C. et al. title Interpretable machine learning: Fundamental principles and 10 grand challenges . journal Statistics Surveys volume 16 , pages 1--85 ( year 2022 )

work page 2022

[15] [15]

& author Armon, A

author Shwartz-Ziv, R. & author Armon, A. title Tabular data: Deep learning is not all you need . journal Information Fusion volume 81 , pages 84--90 ( year 2022 )

work page 2022

[16] [16]

, author Oyallon, E

author Grinsztajn, L. , author Oyallon, E. & author Varoquaux, G. title Why do tree-based models still outperform deep learning on typical tabular data? journal Advances in Neural Information Processing Systems volume 35 , pages 507--520 ( year 2022 )

work page 2022

[17] [17]

& author Maathuis, M

author Drton, M. & author Maathuis, M. H. title Structure learning in graphical modeling . journal Annual Review of Statistics and its Application volume 4 , pages 365--393 ( year 2017 )

work page 2017

[18] [18]

& author Friedman, N

author Koller, D. & author Friedman, N. title Probabilistic Graphical Models: Principles and Techniques ( publisher MIT press , year 2009 )

work page 2009

[19] [19]

author Weinstein, J. N. et al. title The cancer genome atlas pan-cancer analysis project . journal Nature Genetics volume 45 , pages 1113--1120 ( year 2013 )

work page 2013

[20] [20]

, author Hastie, T

author Friedman, J. , author Hastie, T. & author Tibshirani, R. title Sparse inverse covariance estimation with the graphical lasso . journal Biostatistics volume 9 , pages 432--441 ( year 2008 )

work page 2008

[21] [21]

& author B \"u hlmann, P

author Meinshausen, N. & author B \"u hlmann, P. title High-dimensional graphs and variable selection with the lasso . journal The Annals of Statistics volume 34 , pages 1436--1462 ( year 2006 )

work page 2006

[22] [22]

& author Miller, J

author Melikechi, O. & author Miller, J. W. title Integrated path stability selection . journal Journal of the American Statistical Association ( year 2025 )

work page 2025

[23] [23]

, author Dunson, D

author Melikechi, O. , author Dunson, D. B. & author Miller, J. W. title Nonparametric IPSS: fast, flexible feature selection with false discovery control . journal Bioinformatics volume 41 , pages btaf299 ( year 2025 )

work page 2025

[24] [24]

, author Sechidis, K

author Nogueira, S. , author Sechidis, K. & author Brown, G. title On the stability of feature selection algorithms . journal Journal of Machine Learning Research volume 18 , pages 1--54 ( year 2018 )

work page 2018

[25] [25]

author Maitre, L. et al. title Multi-omics signatures of the human early life exposome . journal Nature communications volume 13 , pages 7024 ( year 2022 )

work page 2022

[26] [26]

title Metabolomics and the multi-omics view of cancer

author Wishart, D. title Metabolomics and the multi-omics view of cancer . journal Metabolites volume 12 , pages 154 ( year 2022 )

work page 2022

[27] [27]

author Messer, L. C. , author Jagai, J. S. , author Rappazzo, K. M. & author Lobdell, D. T. title Construction of an environmental quality index for public health research . journal Environmental Health volume 13 , pages 1--22 ( year 2014 )

work page 2014

[28] [28]

, author Egerter, S

author Braveman, P. , author Egerter, S. & author Williams, D. R. title The social determinants of health: coming of age . journal Annual Review of Public Health volume 32 , pages 381--398 ( year 2011 )

work page 2011

[29] [29]

, author Davis, J

author Wei, Y. , author Davis, J. & author Bina, W. F. title Ambient air pollution is associated with the increased incidence of breast cancer in us . journal International Journal of Environmental Health Research volume 22 , pages 12--21 ( year 2012 )

work page 2012

[30] [30]

author Wang, N. et al. title Lung cancer and particulate pollution: A critical review of spatial and temporal analysis evidence . journal Environmental Research volume 164 , pages 585--596 ( year 2018 )

work page 2018

[31] [31]

author Lipfert, F. W. & author Wyzga, R. E. title Longitudinal relationships between lung cancer mortality rates, smoking, and ambient air quality: a comprehensive review and analysis . journal Critical Reviews in Toxicology volume 49 , pages 790--818 ( year 2019 )

work page 2019

[32] [32]

author Kentros, P. A. et al. title Ambient particulate matter air pollution exposure and ovarian cancer incidence in the usa: an ecological study . journal BJOG: An International Journal of Obstetrics & Gynaecology volume 131 , pages 690--698 ( year 2024 )

work page 2024

[33] [33]

author Organization, W. H. title WHO global air quality guidelines: Particulate matter (PM _ 2.5 and PM _ 10 ), ozone, nitrogen dioxide, sulfur dioxide and carbon monoxide ( publisher World Health Organization , year 2021 )

work page 2021

[34] [34]

, author Zhao, C

author Luo, Y. , author Zhao, C. & author Chen, F. title Multiomics research: Principles and challenges in integrated analysis . journal BioDesign Research volume 6 , pages 0059 ( year 2024 )

work page 2024

[35] [35]

, author Seldin, M

author Hasin, Y. , author Seldin, M. & author Lusis, A. title Multi-omics approaches to disease . journal Genome Biology volume 18 , pages 1--15 ( year 2017 )

work page 2017

[36] [36]

author Hayes, C. N. , author Nakahara, H. , author Ono, A. , author Tsuge, M. & author Oka, S. title From omics to multi-omics: A review of advantages and tradeoffs . journal Genes volume 15 , pages 1551 ( year 2024 )

work page 2024

[37] [37]

author Wekesa, J. S. & author Kimwele, M. title A review of multi-omics data integration through deep learning approaches for disease diagnosis, prognosis, and treatment . journal Frontiers in Genetics volume 14 , pages 1199087 ( year 2023 )

work page 2023

[38] [38]

author Ganini, C. et al. title Global mapping of cancers: The Cancer Genome Atlas and beyond . journal Molecular Oncology volume 15 , pages 2823--2840 ( year 2021 )

work page 2021

[39] [39]

author Mallick, H. et al. title An integrated bayesian framework for multi-omics prediction and classification . journal Statistics in Medicine volume 43 , pages 983--1002 ( year 2024 )

work page 2024

[40] [40]

author Mohr, A. E. , author Ortega-Santos, C. P. , author Whisner, C. M. , author Klein-Seetharaman, J. & author Jasbi, P. title Navigating challenges and opportunities in multi-omics integration for personalized healthcare . journal Biomedicines volume 12 , pages 1496 ( year 2024 )

work page 2024

[41] [41]

author Christgen, M. et al. title Lobular breast cancer: Clinical, molecular and morphological characteristics . journal Pathology-Research and Practice volume 212 , pages 583--597 ( year 2016 )

work page 2016

[42] [42]

title The prognosis of invasive ductal carcinoma, lobular carcinoma and mixed ductal and lobular carcinoma according to molecular subtypes of the breast

author Zhao, H. title The prognosis of invasive ductal carcinoma, lobular carcinoma and mixed ductal and lobular carcinoma according to molecular subtypes of the breast . journal Breast Cancer volume 28 , pages 187--195 ( year 2021 )

work page 2021

[43] [43]

& author Metzger-Filho, O

author Barroso-Sousa, R. & author Metzger-Filho, O. title Differences between invasive lobular and invasive ductal carcinoma of the breast: results and therapeutic implications . journal Therapeutic Advances in Medical Oncology volume 8 , pages 261--266 ( year 2016 )

work page 2016

[44] [44]

, author Luo, J.-H

author Zhang, X.-X. , author Luo, J.-H. & author Wu, L.-Q. title FN1 overexpression is correlated with unfavorable prognosis and immune infiltrates in breast cancer . journal Frontiers in Genetics volume 13 , pages 913659 ( year 2022 )

work page 2022

[45] [45]

author Noman, M. Z. et al. title Tumor-promoting effects of myeloid-derived suppressor cells are potentiated by hypoxia-induced expression of miR-210 . journal Cancer Research volume 75 , pages 3771--3787 ( year 2015 )

work page 2015

[46] [46]

author McCormick, R. et al. title miR-210 is a target of hypoxia-inducible factors 1 and 2 in renal cancer, regulates ISCU and correlates with good prognosis . journal British Journal of Cancer volume 108 , pages 1133--1142 ( year 2013 )

work page 2013

[47] [47]

author Yang, F. et al. title MiR-210 in exosomes derived from CAFs promotes non-small cell lung cancer migration and invasion through PTEN/PI3K/AKT pathway . journal Cellular Signalling volume 73 , pages 109675 ( year 2020 )

work page 2020

[48] [48]

author Kojima, S. et al. title Tumour suppressors miR-1 and miR-133a target the oncogenic function of purine nucleoside phosphorylase (PNP) in prostate cancer . journal British Journal of Cancer volume 106 , pages 405--413 ( year 2012 )

work page 2012

[49] [49]

& author Lilly, B

author Sawant, D. & author Lilly, B. title MicroRNA-145 targets in cancer and the cardiovascular system: evidence for common signaling pathways . journal Vascular Biology volume 2 , pages R115--R128 ( year 2020 )

work page 2020

[50] [50]

author Sharma, N. et al. title BCR/ABL1 and BCR are under the transcriptional control of the MYC oncogene . journal Molecular Cancer volume 14 , pages 1--11 ( year 2015 )

work page 2015

[51] [51]

, author Stern, A

author Yang, H.-C. , author Stern, A. & author Chiu, D. T.-Y. title G6PD: A hub for metabolic reprogramming and redox signaling in cancer . journal Biomedical Journal volume 44 , pages 285--292 ( year 2021 )

work page 2021

[52] [52]

& author Hastie, T

author Mazumder, R. & author Hastie, T. title Exact covariance thresholding into connected components for large-scale graphical lasso . journal The Journal of Machine Learning Research volume 13 , pages 781--794 ( year 2012 )

work page 2012

[53] [53]

, author Lafferty, J

author Liu, H. , author Lafferty, J. & author Wasserman, L. title The nonparanormal: S emiparametric estimation of high dimensional undirected graphs. journal Journal of Machine Learning Research volume 10 ( year 2009 )

work page 2009

[54] [54]

& author Zou, H

author Xue, L. & author Zou, H. title Regularized rank-based estimation of high-dimensional nonparanormal graphical models . journal The Annals of Statistics volume 40 , pages 2541--2571 ( year 2012 )

work page 2012

[55] [55]

title Gaussian graphical model estimation with false discovery rate control

author Liu, W. title Gaussian graphical model estimation with false discovery rate control . journal The Annals of Statistics volume 41 , pages 2948--2978 ( year 2013 )

work page 2013

[56] [56]

, author Ren, Z

author Zhang, R. , author Ren, Z. & author Chen, W. title Silggm: An extensive r package for efficient statistical inference in large-scale gene networks . journal PLoS Computational Biology volume 14 , pages e1006369 ( year 2018 )

work page 2018

[57] [57]

& author Maathuis, M

author Li, J. & author Maathuis, M. H. title GGM knockoff filter: False discovery rate control for Gaussian graphical models . journal Journal of the Royal Statistical Society Series B: Statistical Methodology volume 83 , pages 534--558 ( year 2021 )

work page 2021

[58] [58]

author Storey, J. D. title The positive false discovery rate: A B ayesian interpretation and the q-value . journal The Annals of Statistics volume 31 , pages 2013--2035 ( year 2003 )

work page 2013

[59] [59]

author Vasaikar, S. V. , author Straub, P. , author Wang, J. & author Zhang, B. title Linkedomics: analyzing multi-omics data within and across 32 cancer types . journal Nucleic Acids Research volume 46 , pages D956--D963 ( year 2018 )

work page 2018

[60] [60]

& author Guestrin, C

author Chen, T. & author Guestrin, C. title Xgboost: A scalable tree boosting system . In booktitle Proceedings of the 22nd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining , pages 785--794 ( year 2016 )

work page 2016

[61] [61]

, author Liu, H

author Zhao, T. , author Liu, H. , author Roeder, K. , author Lafferty, J. & author Wasserman, L. title The huge package for high-dimensional undirected graph estimation in r . journal The Journal of Machine Learning Research volume 13 , pages 1059--1062 ( year 2012 )

work page 2012

[62] [62]

, author Sun, T

author Ren, Z. , author Sun, T. , author Zhang, C.-H. & author Zhou, H. H. title Asymptotic normality and optimalities in estimation of large Gaussian graphical models . journal The Annals of Statistics volume 43 , pages 991--1026 ( year 2015 )

work page 2015

[63] [63]

& author Van De Geer, S

author Jankova, J. & author Van De Geer, S. title Confidence intervals for high-dimensional inverse covariance estimation . journal Electronic Journal of Statistics volume 9 , pages 1205--1229 ( year 2015 )

work page 2015

[64] [64]

& author van de Geer, S

author Jankov \'a , J. & author van de Geer, S. title Honest confidence regions and optimality in high-dimensional precision matrix estimation . journal Test volume 26 , pages 143--162 ( year 2017 )

work page 2017