Pith. sign in

REVIEW 4 major objections 5 minor 63 references

Annealed variational mixtures for disease subtyping and biomarker discovery

T0 review · 4 major / 5 minor · reviewed 2026-08-12 · deepseek-v4-flash

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

Pith's one-line read The paper claims that VBVarSel, an annealed variational Bayes algorithm, simultaneously clusters high-dimensional samples and selects the biomarkers that define the clusters, outperforming current methods in both speed and accuracy.

desk verdict A practical annealed-VI + variable-selection combination that is genuinely new, but 'outperforms' should read 'matches accuracy with a big speedup,' and the diagonal-covariance assumption is untested where real omics data live. read the letter →

arxiv 2411.19262 v1 pith:Y2L7HI7C submitted 2024-11-28 stat.CO stat.AP

Emma Prevot , Rory Toogood , Filippo Pagani , Paul D. W. Kirk This is my paper · ORCID
classification stat.COstat.AP MSC 62F1562H3062P10
keywords variationalBayesannealedinferencemixturemodelsvariableselectioncancersubtypingbiomarkerdiscoveryhigh-dimensionalclusteringoverfittedmixtures
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

The paper introduces VBVarSel, an annealed variational Bayes algorithm that fits high-dimensional Gaussian mixture models while simultaneously deciding which variables are informative for clustering and which are noise. The authors' central claim is that VBVarSel is both faster and more accurate than existing tools for this joint task, and they support it with simulations and two cancer datasets: breast-cancer gene expression and a pan-cancer protein panel. If the claim holds, disease subtyping and biomarker discovery could be run on full, unfiltered omics data without first reducing dimension or pre-selecting genes. The practical payoff is one open-source Python routine that returns cluster labels, a set of candidate biomarkers, and an inferred number of subtypes in minutes to about an hour.

What carries the argument

The engine is an annealed mean-field variational Bayes routine over an overfitted finite Gaussian mixture with a diagonal per-cluster precision matrix. Three moving parts do the work: a temperature parameter multiplying the entropy term in the ELBO, which flattens the objective at high temperature and is reduced over time so the optimizer escapes shallow local optima; a latent binary indicator per variable that switches each covariate between its cluster-specific distribution and a common null distribution, so variable selection and clustering are solved in one objective; and a symmetric Dirichlet prior with concentration below one that lets surplus mixture components empty out, so the number of clusters is inferred rather than fixed. All conditional updates stay conjugate, so each iteration is a set of closed-form coordinate updates.

What would settle it

Take a dataset with known cluster labels and a strong within-cluster correlation block, such as 200 simulated features in three clusters with pairwise correlation above 0.8 inside a co-expressed module plus 20 pure signal features, run VBVarSel, and compare its Adjusted Rand Index and variable-selection accuracy against ground truth; if either metric falls far below the near-perfect values reported for independent features, the central robustness claim is contradicted. The TCGA PAM50 analysis offers a real-data check: force the model to include correlated neighbours of the PAM50 genes and see whether PAM50 enrichment remains far above random.

Watch

Extended reading notes

Core claim

The central claim is that combining variational inference, simulated-annealing temperature schedules, and latent binary feature-selection indicators makes simultaneous clustering and variable selection scalable enough for real biomedical data while matching or exceeding the accuracy of slower reference methods. On the Crook et al. simulation design, VBVarSel recovers the true three-cluster structure and the exact relevant-variable set in most settings, with runtimes of about 1 to 28 seconds for 100 to 1000 observations and 200 variables, against minutes to tens of thousands of seconds for competing implementations. On a 348 by 17373 TCGA breast-cancer transcriptomic matrix it converges in under an hour without preprocessing, selects roughly 39% of genes, and retains the PAM50 marker genes at a rate the paper reports as far above random and consistent with known breast-cancer subtypes. On a 5157 by 217 TCPA protein panel it produces clusters that largely match tissue of origin and groups related cancers together, keeping about 90% of the pre-selected proteins. The paper concludes that VBVarSel outperforms the current state of the art in accuracy and runtime, and credits the annealing component for improved robustness to poor initialization, correlated features, and noise.

Load-bearing premise

The claim's load-bearing premise is that, within each cluster, the measured features are independent of one another, because the model uses a diagonal precision matrix and the paper only tests correlations up to 0.5 in simulation; if real omics data carry stronger or structured correlation, the cluster labels and biomarker lists could be systematically biased.

Editorial extensions

If this is right

  • On the paper's simulations, VBVarSel recovers the true clustering and the relevant-variable set at near-perfect rates while running at least 2.5 times faster than the next-fastest method at 1000 observations, and it can complete the full TCGA breast-cancer study in under an hour.
  • The method removes the need for a separate feature-selection step: cluster assignment and biomarker identification come from the same fitted model, and the overfitted-mixture prior makes a pre-specified number of clusters unnecessary.
  • Annealing is what buys robustness: the paper reports that geometric or harmonic temperature schedules restore near-perfect accuracy under sub-optimal initialization, added Gaussian noise, and within-cluster correlations up to 0.5, where the non-annealed version degrades.
  • Because the objective is variational, the algorithm produces a full approximate posterior over allocations and selection indicators, not just point estimates, which is what allows direct probabilistic statements about biomarker relevance.
  • The two real-data applications suggest the same code can handle both wide transcriptomic data with 17,373 genes and 348 samples and tall proteomic data with 217 proteins and 5,157 samples, positioning it as a general tool for omics subtyping.

Reading between the lines

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

  • This inference goes beyond the paper: the conditional-independence assumption is the natural boundary of the speed advantage, and on data with strong co-expression modules or long-range correlation, a sparse-precision or factor-analytic covariance variant would likely be needed.
  • This inference goes beyond the paper: the reported sensitivity to the scale hyperparameter and the empirical choice of temperature schedule suggest that automatic schedule and hyperparameter selection would be the highest-value usability improvement, though the paper does not propose one.
  • This inference goes beyond the paper: the binary selection indicator may be why the TCPA analysis retains 90% of variables, and a continuous saliency score could better capture subtle biomarker importance, which the authors mention as future work.
  • This inference goes beyond the paper: a direct scaling test on single-cell RNA-seq or multi-omics matrices would show whether the runtime advantage generalizes, since both real applications in the paper are bulk tissue with standard array technologies.
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 introduces VBVarSel, an annealed variational Bayes algorithm for simultaneous clustering and variable selection in high-dimensional Gaussian mixture models. The model assumes conditional independence of covariates within clusters (diagonal precision matrices), uses an overfitted mixture with a Dirichlet prior on weights to infer the number of clusters, and models variable selection through Bernoulli indicators with Beta priors. The authors derive mean-field variational updates, add an annealing temperature schedule, and validate the method on simulations following Crook et al. (2019), comparing runtime, variable-selection accuracy, and Adjusted Rand Index against hierarchical clustering, sparse K-means, SUGSVarSel, VSCC, and VarSelLCM. They also apply VBVarSel to TCGA breast cancer transcriptomics and TCPA proteomics data. The central claims are that VBVarSel is scalable, computationally efficient, and outperforms the current state of the art in accuracy and runtime.

Significance. If fully substantiated, VBVarSel would be a practically valuable addition to the toolkit for high-dimensional biomedical subtyping: it provides an open-source Python implementation, handles feature selection and cluster-number selection jointly, and shows large runtime advantages over MCMC- and IC-based alternatives in the reported experiments. The manuscript deserves credit for reporting quartiles over repeated simulation runs, for benchmarking on an externally generated simulation design, and for making code publicly available. However, the evidence as presented supports runtime advantages more strongly than accuracy advantages, and the central 'outperforms the state of the art' claim rests on settings in which the model's conditional-independence assumption is either exactly satisfied or only mildly violated.

major comments (4)
  1. [Eq. (2) / Supplement Eq. (18), Supplement Tables 2-4] The model's central factorization assumes that covariates are independent given cluster allocation, so that each within-cluster precision matrix is diagonal. The robustness experiments in Supplement Tables 2-4 only introduce correlations up to 0.5, and the main simulation benchmark uses exactly diagonal covariances. Real transcriptomic and proteomic data commonly exhibit block-structured correlation with high within-block values (0.7-0.95), under which the variational variable-selection indicators and cluster responsibilities can be distorted. Please add simulation scenarios with block-diagonal within-cluster covariance at realistic correlation levels and report ARI, selection proportions, and runtime, or explicitly restrict the claim of state-of-the-art accuracy to near-diagonal settings. This is load-bearing for the abstract and Discussion claims about biomedical applications.
  2. [Results, Simulation study, Tables 1-4] The simulation results do not support an accuracy advantage over VarSelLCM: in Tables 1-4, VarSelLCM matches VBVarSel's perfect or near-perfect ARI and variable-selection proportions in every setting, and SUGSVarSel also attains perfect results in the n=1000 rows of Tables 3 and 4. The text's statement that 'VBVarSel and VarSelLCM are the only two methods that consistently achieve perfect results' is therefore inaccurate. The supported advantage of VBVarSel over these methods is runtime, not accuracy. Please revise the accuracy claims and, ideally, benchmark against an additional recent method on real data where ground-truth subtypes are available for quantitative comparison.
  3. [Algorithm 1; Eqs. (9) and (62); Supplement Section 1.5] Algorithm 1 monitors the standard ELBO of Eq. (62) for convergence, but for T>1 the quantity being optimized is the annealed ELBO of Eq. (9), which does not provide a lower bound on the standard ELBO. Under fixed T>1 or during the early phase of a geometric/harmonic schedule, the standard ELBO need not be monotone, so the convergence criterion '0 < improve < epsilon' may fail to trigger or may stop at an iterate that does not optimize the annealed objective. Please state explicitly which objective is monitored, use the annealed objective for convergence and model selection, or provide a monotonicity result for the annealed updates.
  4. [Supplement Section 3.1; Discussion] The manuscript itself reports that 'VBVarSel requires very careful tuning of b0j' and that 'even slight deviations from optimal would significantly and detrimentally impact the quality of the stratification', and the Discussion states that performance is sensitive to the temperature schedule. Given that the paper's selling point is a scalable, efficient, turnkey alternative to existing methods, this acknowledged sensitivity needs systematic treatment: report a sensitivity analysis over b0j, initial temperature, and schedule type, and provide explicit practical guidance on how these hyperparameters should be chosen in new applications.
minor comments (5)
  1. [Abstract] The abstract contains a typo: 'biomedeical' should be 'biomedical'.
  2. [Table 1, SUGSVarSel row] In Table 1, the SUGSVarSel runtime entry for n=100 is reported as '9.71 [9.98, 8.85]', with the median lying below the lower quartile and above the upper quartile; the quartile order appears to be reversed or the entries are mislabeled.
  3. [Supplement Eq. (37)] In Eq. (37), the second term inside the sum should be a log-density, i.e. (1-c_j) ln f_j(x_nj | Phi_0j), not (1-c_j) f_j(x_nj | Phi_0j), as written.
  4. [Supplement Eqs. (66)-(67)] The expression in Eq. (67) contains '2v' in the digamma argument, which should presumably be '2d0' as in Eq. (66); please correct the notation.
  5. [Performance evaluation section] The description of model selection says that multiple parameter initializations and temperature schedules are evaluated and the one maximizing the ELBO is chosen; because this selection is done on the same data later used to report performance, the manuscript should clarify whether any held-out or repeated-splitting procedure is used, or note that this constitutes tuning on the evaluation set.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the variational updates and simulation benchmarks are self-contained, with author-overlapping citations used only as comparators, not as justification.

full rationale

This paper's claimed derivation chain is the annealed mean-field variational update for a diagonal-covariance Gaussian mixture with binary feature-selection indicators (Supplementary Eqs. 12-73). Each update (responsibilities r_nk, indicator expectations c_j, and Gaussian-Gamma parameters) is obtained by the standard coordinate-ascent stationarity condition on the annealed ELBO; none of the update equations assumes the reported outcomes (ARI, selection proportions, runtimes). The simulation benchmark is inherited from Crook et al. (2019), which includes a co-author (Kirk), but it is used as an external data generator and as a comparator, not as the justification for VBVarSel's correctness; the paper's own tables provide the empirical comparison. The temperature schedule is chosen empirically by maximizing the ELBO, and the paper explicitly acknowledges sensitivity to this choice in the Discussion: "its performance is still sensitive to the choice of temperature schedule. Careful tuning of this parameter is required." Selecting hyperparameters by ELBO is model selection within the same inference objective; the target metrics (ARI versus ground-truth labels, selection of permuted covariates) are not functions of the ELBO and are not forced by that selection. No specific equation reduces to its own input, and no load-bearing claim rests solely on a self-citation. Therefore no significant circularity is present.

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

The central performance claim rests on a set of user-chosen hyperparameters (prior scales and the annealing schedule) plus standard VI assumptions. The heavy reliance on the user to tune b0j and the temperature schedule is the most significant unresolved burden; the paper provides recommended ranges but no automatic tuning.

free parameters (5)
  • b0j = [0.1, 1] for synthetic and TCGA; 0.1 for TCPA
    Scale parameter of the Gamma prior on the cluster precisions. Supplement Section 3.1 states: 'VBVarSel requires very careful tuning of b0j. Even slight deviations from optimal would significantly and detrimentally impact the quality of the stratification.' This is a hand-chosen hyperparameter with strong effect.
  • alpha0 = [0.1, 1] synthetic; [0.01, 0.1] TCGA; 1/K for A-TCGA; 10^-3 for TCPA
    Dirichlet concentration on mixture weights. Controls how strongly extra clusters are emptied; values below 0.5 worked for simulations, but the choice is data-dependent.
  • d0 = 0.9 synthetic; [1,5] or [0.9,5] for TCGA; 0.5 for TCPA
    Shape of the Beta prior on variable-selection probabilities. Values below 0.5 lead to higher deselection, above 5 to lower deselection; the paper uses different values per dataset.
  • temperature schedule (T0, schedule type, annealed iterations) = T = 2, 3, or 4; geometric/harmonic/fixed; 5-10 annealed iterations
    The annealing schedule is chosen empirically; the paper states 'since there is no consensus in the literature on the optimal type of schedule, we determine our approach empirically.' For some experiments they report only the schedules that 'allowed more significant advantages.' This is a free modeling choice.
  • K (maximum number of clusters) = [3,10] synthetic; [5,8] TCGA; [25,40] TCPA
    The overfitted mixture is initialized with a large K, and the range is dataset-specific.
assumptions (5)
  • domain assumption Within each cluster, covariates are conditionally independent (diagonal precision matrix).
    Main text Eq. (2) and Supplement Eq. (18): f_x(x_n|Φ_k) = ∏_j f_j(x_nj|Φ_kj), equivalent to Λ_k diagonal. This excludes correlated features within a subtype, a simplification for real omics data.
  • domain assumption Overfitted mixtures with α0 < 1 asymptotically shrink extra components to zero.
    Adopted from Rousseau and Mengersen (2011); the paper relies on this to infer the number of clusters from an overfitted K. The finite-sample validity in the variational setting is taken as given.
  • standard math Mean-field factorization of the variational distribution.
    Eq. (7): q(θ) = ∏_i q_i(θ_i). Standard mean-field VI assumption; the optimization is convex in each factor.
  • standard math Conjugate priors (Gaussian-Gamma, Dirichlet, Beta-Bernoulli) yield tractable updates.
    Priors specified in Section 1.2 of the supplement; conjugacy is used to derive closed-form update equations.
  • domain assumption The annealed ELBO with T>1 flattens the objective and helps escape local optima.
    The paper relies on the statistical-mechanics intuition (Rose et al. 1990, Katahira et al. 2008) that annealing improves optimization; no guarantee is proven for this specific model.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Annealed variational mixtures for disease subtyping and biomarker discovery." pith.science (2026). https://pith.science/paper/Y2L7HI7C

@misc{pith2026241119262,
  author       = {Pith},
  title        = {Pith review of: Annealed variational mixtures for disease subtyping and biomarker discovery},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/Y2L7HI7C}},
  note         = {Machine review of arXiv:2411.19262}
}
read the original abstract

Cluster analyses of high-dimensional data are often hampered by the presence of large numbers of variables that do not provide relevant information, as well as the perennial issue of choosing an appropriate number of clusters. These challenges are frequently encountered when analysing `omics datasets, such as in molecular precision medicine, where a key goal is to identify disease subtypes and the biomarkers that define them. Here we introduce an annealed variational Bayes algorithm for fitting high-dimensional mixture models while performing variable selection. Our algorithm is scalable and computationally efficient, and we provide an open source Python implementation, VBVarSel. In a range of simulated and real biomedical examples, we show that VBVarSel outperforms the current state of the art, and demonstrate its use for cancer subtyping and biomarker discovery.

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

63 extracted references · 51 canonical work pages

  1. [1]

    Nature communications 5(1):3887

    Akbani R, Ng PKS, Werner HM, et al (2014) A pan-cancer proteomic perspective on the cancer genome atlas. Nature communications 5(1):3887

    work page 2014

  2. [2]

    Journal of Classification 31(2):136--153

    Andrews JL, McNicholas PD (2014) Variable selection for clustering and classification. Journal of Classification 31(2):136--153

    work page 2014

  3. [3]

    Princeton University Press

    Bellman R (1957) Dynamic Programming . Princeton University Press

    work page 1957

  4. [4]

    In: Rizzi A, Vichi M, Bock HH (eds) Advances in Data Science and Classification

    Bensmail H, Meulman JJ (1998) MCMC inference for model-based cluster analysis. In: Rizzi A, Vichi M, Bock HH (eds) Advances in Data Science and Classification. Springer Berlin Heidelberg, pp 191--196

    work page 1998

  5. [5]

    Springer-Verlag, Berlin, Heidelberg

    Bishop CM (2006) Pattern Recognition and Machine Learning (Information Science and Statistics). Springer-Verlag, Berlin, Heidelberg

    work page 2006

  6. [6]

    Journal of the American Statistical Association 112(518):859--877

    Blei DM, Kucukelbir A, McAuliffe JD (2017) Variational inference: A review for statisticians. Journal of the American Statistical Association 112(518):859--877. doi:10.1080/01621459.2017.1285773

    arXiv 2017

  7. [7]

    Computational Statistics & Data Analysis 71:52--78

    Bouveyron C, Brunet-Saumard C (2014) Model-based clustering of high-dimensional data: A review. Computational Statistics & Data Analysis 71:52--78. doi:https://doi.org/10.1016/j.csda.2012.12.008

    doi:10.1016/j.csda.2012.12.008 2014

  8. [8]

    Cambridge university press

    Boyd SP, Vandenberghe L (2004) Convex optimization. Cambridge university press

    work page 2004

Show all 63 references
  1. [9]

    Journal de la Societe francaise de statistique (2009) 155

    Celeux G, Martin-Magniette ML, Maugis-Rabusseau C, et al (2013) Comparing model selection and regularization approaches to variable selection in model-based clustering. Journal de la Societe francaise de statistique (2009) 155

    2013

  2. [10]

    IEEE Transactions on Pattern Analysis and Machine Intelligence 28(6):1013--1018

    Constantinopoulos C, Titsias MK, Likas A (2006) Bayesian feature and model selection for gaussian mixture models. IEEE Transactions on Pattern Analysis and Machine Intelligence 28(6):1013--1018. doi:10.1109/TPAMI.2006.111

    2006 doi

  3. [11]

    Current Research in Biotechnology 4:138--151

    Cremin CJ, Dash S, Huang X (2022) Big data: Historic advances and emerging trends in biomedical research. Current Research in Biotechnology 4:138--151. doi:https://doi.org/10.1016/j.crbiot.2022.02.004, ://www.sciencedirect.com/science/article/pii/S2590262822000090

    2022 doi

  4. [12]

    Statistical Applications in Genetics and Molecular Biology 18(6):20180065

    Crook OM, Gatto L, Kirk PDW (2019) Fast approximate inference for variable selection in dirichlet process mixtures, with an application to pan-cancer proteomics. Statistical Applications in Genetics and Molecular Biology 18(6):20180065. doi:doi:10.1515/sagmb-2018-0065

    2019 doi

  5. [13]

    Nature 486(7403):346--352

    Curtis C, Shah SP, Chin SF, et al (2012) The genomic and transcriptomic architecture of 2,000 breast tumours reveals novel subgroups. Nature 486(7403):346--352

    2012

  6. [14]

    CPT: pharmacometrics & systems pharmacology 2(3):1--10

    Duan Q, Kou Y, Clark N, et al (2013) Metasignatures identify two major subtypes of breast cancer. CPT: pharmacometrics & systems pharmacology 2(3):1--10

    2013

  7. [15]

    Physical Chemistry Chemical Physics 7(23):3910--3916

    Earl DJ, Deem MW (2005) Parallel tempering: Theory, applications, and new perspectives. Physical Chemistry Chemical Physics 7(23):3910--3916

    2005

  8. [16]

    Statistics Surveys 12:18 -- 65

    Fop M, Murphy TB (2018) Variable selection methods for model-based clustering . Statistics Surveys 12:18 -- 65. doi:10.1214/18-SS119

    2018 doi

  9. [17]

    Journal of the American Statistical Association 97(458):611--631

    Fraley C, Raftery AE (2002) Model-based clustering, discriminant analysis, and density estimation. Journal of the American Statistical Association 97(458):611--631. doi:10.1198/016214502760047131

    2002 doi

  10. [18]

    Journal of Classification 12:113–136

    Gnanadesikan R, Kettenring J, Tsao S (1995) Weighting and selection of variables for cluster analysis. Journal of Classification 12:113–136. doi:https://doi.org/10.1007/BF01202271

    1995 doi

  11. [19]

    science 286(5439):531--537

    Golub TR, Slonim DK, Tamayo P, et al (1999) Molecular classification of cancer: class discovery and class prediction by gene expression monitoring. science 286(5439):531--537

    1999

  12. [20]

    Artificial Intelligence Review 53:4519--4545

    Hancer E, Zhang M, Xue B (2020) A survey on feature selection approaches for clustering. Artificial Intelligence Review 53:4519--4545. doi:https://doi.org/10.1007/s10462-019-09800-w

    2020 doi

  13. [21]

    Math Intell 27:83--85

    Hastie T, Tibshirani R, Friedman J, et al (2004) The elements of statistical learning: Data mining, inference, and prediction. Math Intell 27:83--85. doi:10.1007/BF02985802

    2004 doi

  14. [22]

    Cell 158(4):929--944

    Hoadley KA, Yau C, Wolf DM, et al (2014) Multiplatform analysis of 12 cancer types reveals molecular classification within and across tissues of origin. Cell 158(4):929--944

    2014

  15. [23]

    Journal of Classification 2:193--218

    Hubert LJ, Arabie P (1985) Comparing partitions. Journal of Classification 2:193--218

    1985

  16. [24]

    Journal of Physics: Conference Series 95(1):012015

    Katahira K, Watanabe K, Okada M (2008) Deterministic annealing variant of variational bayes method. Journal of Physics: Conference Series 95(1):012015. doi:10.1088/1742-6596/95/1/012015, ://dx.doi.org/10.1088/1742-6596/95/1/012015

    2008 doi

  17. [25]

    2303.00318

    Kirk PDW, Pagani F, Richardson S (2023) Bayesian outcome-guided multi-view mixture models with applications in molecular precision medicine. 2303.00318

    2023 arXiv

  18. [26]

    science 220(4598):671--680

    Kirkpatrick S, Gelatt Jr CD, Vecchi MP (1983) Optimization by simulated annealing. science 220(4598):671--680

    1983

  19. [27]

    The annals of mathematical statistics 22(1):79--86

    Kullback S, Leibler RA (1951) On information and sufficiency. The annals of mathematical statistics 22(1):79--86

    1951

  20. [28]

    Journal of Computational and Graphical Statistics - J COMPUT GRAPH STAT 16

    Lau J, Green P (2007) Bayesian model-based clustering procedures. Journal of Computational and Graphical Statistics - J COMPUT GRAPH STAT 16. doi:10.1198/106186007X238855

    2007 doi

  21. [29]

    IEEE Transactions on Pattern Analysis and Machine Intelligence 26(9):1154--1166

    Law MH, Figueiredo MA, Jain AK (2004) Simultaneous feature selection and clustering using mixture models. IEEE Transactions on Pattern Analysis and Machine Intelligence 26(9):1154--1166. doi:10.1109/TPAMI.2004.71

    2004 doi

  22. [30]

    Nature Methods 10:1046--1047

    Li J, Lu Y, Akbani R, et al (2013) TCPA : a resource for cancer functional proteomics data. Nature Methods 10:1046--1047

    2013

  23. [31]

    1303.2836

    Liverani S, Hastie DI, Azizi L, et al (2014) PReMiuM : An R package for profile regression mixture models using dirichlet processes. 1303.2836

    2014 arXiv

  24. [32]

    Bioinformatics 29(20):2610--2616

    Lock EF, Dunson DB (2013) Bayesian consensus clustering. Bioinformatics 29(20):2610--2616

    2013

  25. [33]

    In: Proceedings of the fifth Berkeley symposium on mathematical statistics and probability, Oakland, CA, USA, pp 281--297

    MacQueen J, et al (1967) Some methods for classification and analysis of multivariate observations. In: Proceedings of the fifth Berkeley symposium on mathematical statistics and probability, Oakland, CA, USA, pp 281--297

    1967

  26. [34]

    1411.1810

    Mandt S, McInerney J, Abrol F, et al (2016) Variational tempering. 1411.1810

    2016 arXiv

  27. [35]

    Statistics and Computing 27:1049--1063

    Marbac M, Sedki M (2017) Variable selection for model-based clustering using the integrated complete-data likelihood. Statistics and Computing 27:1049--1063

    2017

  28. [36]

    Bioinformatics 35(7):1255--1257

    Marbac M, Sedki M (2018) VarSelLCM : an R/C++ package for variable selection in model-based clustering of mixed-data with missing values . Bioinformatics 35(7):1255--1257. doi:10.1093/bioinformatics/bty786

    2018 doi

  29. [37]

    Journal of Classification 33:331--373

    McNicholas PD (2016) Model-based clusterig. Journal of Classification 33:331--373

    2016

  30. [38]

    Procedia Computer Science 91:919--926

    Miao J, Niu L (2016) A survey on feature selection. Procedia Computer Science 91:919--926. doi:https://doi.org/10.1016/j.procs.2016.07.111

    2016 doi

  31. [39]

    Nature 490(7418):61--70

    Network TCGA (2012) Comprehensive molecular portraits of human breast tumours. Nature 490(7418):61--70

    2012

  32. [40]

    ://www.nhs.uk/conditions/breast-cancer/, accessed: 2023-07-06

    NHS (2022) Breast cancer in women. ://www.nhs.uk/conditions/breast-cancer/, accessed: 2023-07-06

    2022

  33. [41]

    Breast Cancer [Internet]

    Orrantia-Borunda E, Anchondo-Nu \ n ez P, Acu \ n a-Aguilar LE, et al (2022) Subtypes of breast cancer. Breast Cancer [Internet]

    2022

  34. [42]

    Physics Letters A 73(3):203--205

    Parisi G (1979) Toward a mean field theory for spin glasses. Physics Letters A 73(3):203--205

    1979

  35. [43]

    Journal of clinical oncology 27(8):1160

    Parker JS, Mullins M, Cheang MC, et al (2009) Supervised risk predictor of breast cancer based on intrinsic subtypes. Journal of clinical oncology 27(8):1160

    2009

  36. [44]

    Breast cancer research 12(5):1--18

    Prat A, Parker JS, Karginova O, et al (2010) Phenotypic and molecular characterization of the claudin-low intrinsic subtype of breast cancer. Breast cancer research 12(5):1--18

    2010

  37. [45]

    Journal of the American Statistical Association 66(336):846--850

    Rand WM (1971) Objective criteria for the evaluation of clustering methods. Journal of the American Statistical Association 66(336):846--850. doi:10.1080/01621459.1971.10482356

    1971

  38. [46]

    Pattern Recognition Letters 11(9):589--594

    Rose K, Gurewitz E, Fox G (1990) A deterministic annealing approach to clustering. Pattern Recognition Letters 11(9):589--594. doi:https://doi.org/10.1016/0167-8655(90)90010-Y

    1990 doi

  39. [47]

    Journal of the Royal Statistical Society Series B: Statistical Methodology 73(5):689--710

    Rousseau J, Mengersen K (2011) Asymptotic behaviour of the posterior distribution in overfitted mixture models. Journal of the Royal Statistical Society Series B: Statistical Methodology 73(5):689--710

    2011

  40. [48]

    The Annals of Applied Statistics 14(2):905 -- 928

    Ruffieux H, Davison AC, Hager J, et al (2020) A global-local approach for detecting hotspots in multiple-response regression . The Annals of Applied Statistics 14(2):905 -- 928. doi:10.1214/20-AOAS1332

    2020 doi

  41. [49]

    Molecular & Cellular Proteomics 4(4):346--355

    Sheehan KM, Calvert VS, Kay EW, et al (2005) Use of reverse phase protein microarrays and reference standard development for molecular network analysis of metastatic ovarian carcinoma. Molecular & Cellular Proteomics 4(4):346--355

    2005

  42. [50]

    Proceedings of the national academy of sciences 100(14):8418--8423

    S rlie T, Tibshirani R, Parker J, et al (2003) Repeated observation of breast tumor subtypes in independent gene expression data sets. Proceedings of the national academy of sciences 100(14):8418--8423

    2003

  43. [51]

    Psychometrika 73(1):125--144

    Steinley D, Brusco MJ (2008) Selection of variables in cluster analysis: An empirical comparison of eight procedures. Psychometrika 73(1):125--144

    2008

  44. [52]

    Journal of the American Statistical Association 100(470):602--617

    Tadesse MG, Sha N, Vannucci M (2005) Bayesian variable selection in clustering high-dimensional data. Journal of the American Statistical Association 100(470):602--617. doi:10.1198/016214504000001565

    2005 doi

  45. [53]

    Neural networks : the official journal of the International Neural Network Society 11(2):271--282

    Ueda N, Nakano R (1998) Deterministic annealing EM algorithm. Neural networks : the official journal of the International Neural Network Society 11(2):271--282. doi:10.1016/s0893-6080(97)00133-0

    1998 doi

  46. [54]

    Science 357(6352):eaan2507

    Uhlen M, Zhang C, Lee S, et al (2017) A pathology atlas of the human cancer transcriptome. Science 357(6352):eaan2507

    2017

  47. [55]

    Journal of Computational and Graphical Statistics 20(1):196--216

    Wang L, Dunson DB (2011) Fast bayesian inference in dirichlet process mixture models. Journal of Computational and Graphical Statistics 20(1):196--216. doi:10.1198/jcgs.2010.07081

    2011 arXiv

  48. [56]

    Cancers 15(15)

    Wang RC, Wang Z (2023) Precision medicine: Disease subtyping and tailored treatment. Cancers 15(15). doi:10.3390/cancers15153837, ://www.mdpi.com/2072-6694/15/15/3837

    2023 doi

  49. [57]

    Journal of the American statistical association 58(301):236--244

    Ward Jr JH (1963) Hierarchical grouping to optimize an objective function. Journal of the American statistical association 58(301):236--244

    1963

  50. [58]

    Nature genetics 45(10):1113--1120

    Weinstein JN, Collisson EA, Mills GB, et al (2013) The cancer genome atlas pan-cancer analysis project. Nature genetics 45(10):1113--1120

    2013

  51. [59]

    Journal of the American Statistical Association 105(490):713--726

    Witten DM, Tibshirani R (2010) A framework for feature selection in clustering. Journal of the American Statistical Association 105(490):713--726

    2010

  52. [60]

    R package version 1(3)

    Witten DM, Tibshirani R (2013) sparcl: Perform sparse hierarchical clustering and sparse k-means clustering. R package version 1(3)

    2013

  53. [61]

    Journal of Computational and Graphical Statistics 23(4):1143--1162

    Zhang X, Nott DJ, Yau C, et al (2014) A sequential algorithm for fast fitting of dirichlet process mixture models. Journal of Computational and Graphical Statistics 23(4):1143--1162. doi:10.1080/10618600.2013.870906

    2014

  54. [62]

    , " * write output.state after.block = add.period write newline

    ENTRY address archive author booktitle chapter doi edition editor eid eprint howpublished institution journal key keywords month note number organization pages publisher school series title type url volume year archivePrefix primaryClass adsurl adsnote version label extra.labe...

  55. [63]

    write newline

    " write newline "" before.all 'output.state := FUNCTION add.period duplicate empty 'skip "." * add.blank if FUNCTION if.digit duplicate "0" = swap duplicate "1" = swap duplicate "2" = swap duplicate "3" = swap duplicate "4" = swap duplicate "5" = swap duplicate "6" = swap dupl...

Pith tools

Reviewed August 12, 2026 · model on record in the stance chip above.