×

Inferring network structure in non-normal and mixed discrete-continuous genomic data. (English) Zbl 1415.62085

Summary: Inferring dependence structure through undirected graphs is crucial for uncovering the major modes of multivariate interaction among high-dimensional genomic markers that are potentially associated with cancer. Traditionally, conditional independence has been studied using sparse Gaussian graphical models for continuous data and sparse Ising models for discrete data. However, there are two clear situations when these approaches are inadequate. The first occurs when the data are continuous but display non-normal marginal behavior such as heavy tails or skewness, rendering an assumption of normality inappropriate. The second occurs when a part of the data is ordinal or discrete (e.g., presence or absence of a mutation) and the other part is continuous (e.g., expression levels of genes or proteins). In this case, the existing Bayesian approaches typically employ a latent variable framework for the discrete part that precludes inferring conditional independence among the data that are actually observed. The current article overcomes these two challenges in a unified framework using Gaussian scale mixtures. Our framework is able to handle continuous data that are not normal and data that are of mixed continuous and discrete nature, while still being able to infer a sparse conditional sign independence structure among the observed data. Extensive performance comparison in simulations with alternative techniques and an analysis of a real cancer genomics data set demonstrate the effectiveness of the proposed approach.

MSC:

62P10 Applications of statistics to biology and medical sciences; meta analysis
92D10 Genetics and epigenetics
05C90 Applications of graph theory
62H12 Estimation in multivariate analysis

Software:

glasso; BayesLogit
PDFBibTeX XMLCite
Full Text: DOI arXiv Link

References:

[1] Albert, J. H. and Chib, S. (1993). Bayesian analysis of binary and polychotomous response data. {\it Journal of the American statistical Association}88, 669-679. · Zbl 0774.62031
[2] Barndorff‐Nielsen, O. E. (1977). Exponentially Decreasing Distributions for the Logarithm of Particle Size. {\it Royal Society of London Proceedings Series A}353, 401-419.
[3] Barndorff‐Nielsen, O. E. (1978). Hyperbolic distributions and distributions on hyperbolae. {\it Scandinavian Journal of Statistics}5, 151-157. · Zbl 0386.60018
[4] Bhadra, A. and Mallick, B. K. (2013). Joint high‐dimensional Bayesian variable and covariance selection with an application to eQTL analysis. {\it Biometrics}69, 447-457. · Zbl 1274.62722
[5] Brennan, C. W., Verhaak, R. G., McKenna, A., Campos, B., Noushmehr, H., Salama, S. et al. (2013). The somatic genomic landscape of glioblastoma. {\it Cell}155, 462-477.
[6] Carvalho, C. M., Massam, H., and West, M. (2007). Simulation of hyper‐inverse Wishart distributions in graphical models. {\it Biometrika}94, 647-659. · Zbl 1135.62011
[7] Chen, S., Witten, D. M., and Shojaie, A. (2015). Selection and estimation for mixed graphical models. {\it Biometrika}102, 47-64. · Zbl 1345.62081
[8] Cheng, J., Levina, E., and Zhu, J. (2013). High‐dimensional mixed graphical models. arXiv preprint arXiv:1304.2810.
[9] Costa, B., Bendinelli, S., Gabelloni, P., Da Pozzo, E., Daniele, S., Scatena, F. et al. (2013). Human glioblastoma multiforme: p53 reactivation by a novel MDM2 inhibitor. {\it PLoS One}8, e72281.
[10] Dawid, A. P. (1981). Some matrix‐variate distribution theory: Notational considerations and a Bayesian application. {\it Biometrika}68, 265-274. · Zbl 0464.62039
[11] Dobra, A. and Lenkoski, A. (2011). Copula Gaussian graphical models and their application to modeling functional disability data. {\it Annals of Applied Statistics}5, 969-993. · Zbl 1232.62046
[12] Feldman, G., Bhadra, A., and Kirshner, S. (2014). Bayesian feature selection in high‐dimensional regression in presence of correlated noise. {\it Stat}3, 258-272.
[13] Finegold, M. and Drton, M. (2011). Robust graphical modeling of gene networks using classical and alternative t‐distributions. {\it Annals of Applied Statistics}5, 1057-1080. · Zbl 1232.62083
[14] Finegold, M. and Drton, M. (2014). Robust Bayesian graphical modeling using dirichlet t ‐distributions. {\it Bayesian Analysis}9, 521-550. · Zbl 1327.62143
[15] Friedman, J., Hastie, T., and Tibshirani, R. (2008). Sparse inverse covariance estimation with the graphical lasso. {\it Biostatistics}9, 432-441. · Zbl 1143.62076
[16] Furnari, F. B., Fenton, T., Bachoo, R. M., Mukasa, A., Stommel, J., Stegh, A., et al. (2007). Malignant astrocytic glioma: genetics, biology, and paths to treatment. {\it Genes & development}21, 2683-2710.
[17] Gray, J. W. and Collins, C. (2000). Genome changes and gene expression in human solid tumors. {\it Carcinogenesis}21, 443-452.
[18] Griffin, J. E. and Brown, P. J. (2010). Inference with normal‐gamma prior distributions in regression problems. {\it Bayesian Analysis}5, 171-188. · Zbl 1330.62128
[19] Kendall, M. G. (1938). A new measure of rank correlation. {\it Biometrika}30, 81-93. · Zbl 0019.13001
[20] Khatri, C. and Rao, C. R. (1976). Characterizations of multivariate normality. I. through independence of some statistics. {\it Journal of Multivariate Analysis}6, 81-94. · Zbl 0367.62059
[21] Kotz, S., Kozubowski, T. J., and Podgrski, K. (2001). The Laplace Distribution and Generalizations. Boston: Birkhäuser. · Zbl 0977.62003
[22] Lauritzen, S. L. (1996). Graphical Models. Oxford: Oxford University Press. · Zbl 0907.62001
[23] Lee, J. D. and Hastie, T. J. (2015). Learning the structure of mixed graphical models. {\it Journal of Computational and Graphical Statistics}24, 230-253.
[24] Liu, H., Han, F., and Zhang, C.‐H. (2012). Transelliptical graphical models. In Bartlett, P. L., Pereira, F. C. N., Burges, C. J. C., Bottou, L., and Weinberger, K. Q., editors, {\it NIPS}, pages 809-817.
[25] Liu, H., Lafferty, J., and Wasserman, L. (2009). The nonparanormal: Semiparametric estimation of high dimensional undirected graphs. {\it Journal of Machine Learning Research}10, 2295-2328. · Zbl 1235.62035
[26] Marko, N. F. and Weil, R. J. (2012). Non‐Gaussian distributions affect identification of expression patterns, functional annotation, and prospective classification in human cancer genomes. {\it PLoS ONE}7, e46935.
[27] Meinhausen, N. and Bühlmann, P. (2006). High‐dimensional graphs and variable selection with the lasso. {\it Annals of Statistics}34, 1436-1462. · Zbl 1113.62082
[28] Mohammadi, A. and Wit, E. C. (2015). BDgraph: Bayesian structure learning of graphs in R. arXiv preprint arXiv:1501. 05108.
[29] Pitt, M., Chan, D., and Kohn, R. (2006). Efficient Bayesian inference for Gaussian copula regression models. {\it Biometrika}93, 537-554. · Zbl 1108.62027
[30] Polson, N. G., Scott, J. G., and Windle, J. (2012). R package BayesLogit.
[31] Polson, N. G., Scott, J. G., and Windle, J. (2013). Bayesian inference for logistic models using Pólya‐Gamma latent variables. {\it Journal of the American Statistical Association}108, 1339-1349. · Zbl 1283.62055
[32] Ravikumar, P., Wainwright, M. J., and Lafferty, J. D. (2010). High‐dimensional Ising model selection using urn:x-wiley:15410420:media:biom12711:biom12711-math-0223‐regularized logistic regression. {\it Annals of Statistics}38, 1287-1319. · Zbl 1189.62115
[33] Shiraishi, S., Tada, K., Nakamura, H., Makino, K., Kochi, M., Saya, H., et al. (2002). Influence of p53 mutations on prognosis of patients with glioblastoma. {\it Cancer}95, 249-257.
[34] TCGA (2008). Comprehensive genomic characterization defines human glioblastoma genes and core pathways. {\it Nature}455, 1061-1068.
[35] West, M. (1987). On scale mixtures of normal distributions. {\it Biometrika}74, 646-648. · Zbl 0648.62015
[36] Wong, F., Carter, C. K., and Kohn, R. (2003). Efficient estimation of covariance selection models. {\it Biometrika}90, 809-830. · Zbl 1436.62346
[37] Yang, Y., Ravimumar, P., Allen, G., and Liu, Z. (2015). On graphical models via univariate exponential family distributions. {\it Journal of Machine Learning Research}16, 3813-3847. · Zbl 1351.62111
This reference list is based on information provided by the publisher or from digital mathematics libraries. Its items are heuristically matched to zbMATH identifiers and may contain data conversion errors. In some cases that data have been complemented/enhanced by data from zbMATH Open. This attempts to reflect the references listed in the original paper as accurately as possible without claiming completeness or a perfect matching.