arXiv subjects

Korbinian Strimmer

Publications and source records attributed to Korbinian Strimmer.

15 recordsLinked to original sources

arXiv2026

From Physics to Statistics: A Simple Route to Exponential Families via Maximum Entropy

Exponential families form the backbone of modern statistics and machine learning, but textbooks seldom derive them from first principles in an accessible way. Although minimal sufficiency and the principle of maximum entropy, originating in physics, provide core motivation, they are often presented as technical and requiring advanced prerequisites. Here, a short, self-contained derivation of exponential families based on maximum entropy is presented that is straightforward to carry out, requires only a modest background in information entropy, and avoids technicalities like constrained optimisation. Two propositions are demonstrated in this fashion: i) exponential families with a general base maximise information entropy with respect to that base subject to fixed expectations of canonical statistics, and ii) exponential families with a uniform base maximise standard information entropy under the same constraints. Maximum entropy therefore provides a principled foundation for exponential families with minimal prerequisites, highlighting the value of teaching entropy in statistics courses.

stat.ME↗

arXiv2018

A whitening approach to probabilistic canonical correlation analysis for omics data integration

Background: Canonical correlation analysis (CCA) is a classic statistical tool for investigating complex multivariate data. Correspondingly, it has found many diverse applications, ranging from molecular biology and medicine to social science and finance. Intriguingly, despite the importance and pervasiveness of CCA, only recently a probabilistic understanding of CCA is developing, moving from an algorithmic to a model-based perspective and enabling its application to large-scale settings. Results: Here, we revisit CCA from the perspective of statistical whitening of random variables and propose a simple yet flexible probabilistic model for CCA in the form of a two-layer latent variable generative model. The advantages of this variant of probabilistic CCA include non-ambiguity of the latent variables, provisions for negative canonical correlations, possibility of non-normal generative variables, as well as ease of interpretation on all levels of the model. In addition, we show that it lends itself to computationally efficient estimation in high-dimensional settings using regularized inference. We test our approach to CCA analysis in simulations and apply it to two omics data sets illustrating the integration of gene expression data, lipid concentrations and methylation levels. Conclusions: Our whitening approach to CCA provides a unifying perspective on CCA, linking together sphering procedures, multivariate regression and corresponding probabilistic generative models. Furthermore, we offer an efficient computer implementation in the "whitening" R package available at https://CRAN.R-project.org/package=whitening .

stat.ME↗

arXiv2016

Optimal whitening and decorrelation

Whitening, or sphering, is a common preprocessing step in statistical analysis to transform random variables to orthogonality. However, due to rotational freedom there are infinitely many possible whitening procedures. Consequently, there is a diverse range of sphering methods in use, for example based on principal component analysis (PCA), Cholesky matrix decomposition and zero-phase component analysis (ZCA), among others. Here we provide an overview of the underlying theory and discuss five natural whitening procedures. Subsequently, we demonstrate that investigating the cross-covariance and the cross-correlation matrix between sphered and original variables allows to break the rotational invariance and to identify optimal whitening transformations. As a result we recommend two particular approaches: ZCA-cor whitening to produce sphered variables that are maximally similar to the original variables, and PCA-cor whitening to obtain sphered variables that maximally compress the original variables.

stat.ME↗

arXiv2016

Mass spectrometry analysis using MALDIquant

MALDIquant and associated R packages provide a versatile and completely free open-source platform for analyzing 2D mass spectrometry data as generated for instance by MALDI and SELDI instruments. We first describe the various methods and algorithms available in MALDIquant . Subsequently, we illustrate a typical analysis workflow using MALDIquant by investigating an experimental cancer data set, starting from raw mass spectrometry measurements and ending at multivariate classification.

q-bio.QM↗

arXiv2015

Differential protein expression and peak selection in mass spectrometry data by binary discriminant analysis

Motivation: Proteomic mass spectrometry analysis is becoming routine in clinical diagnostics, for example to monitor cancer biomarkers using blood samples. However, differential proteomics and identification of peaks relevant for class separation remains challenging. Results: Here, we introduce a simple yet effective approach for identifying differentially expressed proteins using binary discriminant analysis. This approach works by data-adaptive thresholding of protein expression values and subsequent ranking of the dichotomized features using a relative entropy measure. Our framework may be viewed as a generalization of the `peak probability contrast' approach of Tibshirani et al. (2004) and can be applied both in the two-group and the multi-group setting. Our approach is computationally inexpensive and shows in the analysis of a large-scale drug discovery test data set equivalent prediction accuracy as a random forest. Furthermore, we were able to identify in the analysis of mass spectrometry data from a pancreas cancer study biological relevant and statistically predictive marker peaks unrecognized in the original study. Availability: The methodology for binary discriminant analysis is implemented in the R package binda, which is freely available under the GNU General Public License (version 3 or later) from CRAN at URL http://cran.r-project.org/web/packages/binda/ . R scripts reproducing all described analyzes are available from the web page http://strimmerlab.org/software/binda/ .

q-bio.GN↗

arXiv2015

A Simple Data-Adaptive Probabilistic Variant Calling Model

Background: Several sources of noise obfuscate the identification of single nucleotide variation (SNV) in next generation sequencing data. For instance, errors may be introduced during library construction and sequencing steps. In addition, the reference genome and the algorithms used for the alignment of the reads are further critical factors determining the efficacy of variant calling methods. It is crucial to account for these factors in individual sequencing experiments. Results: We introduce a simple data-adaptive model for variant calling. This model automatically adjusts to specific factors such as alignment errors. To achieve this, several characteristics are sampled from sites with low mismatch rates, and these are used to estimate empirical log-likelihoods. These likelihoods are then combined to a score that typically gives rise to a mixture distribution. From these we determine a decision threshold to separate potentially variant sites from the noisy background. Conclusions: In simulations we show that our simple proposed model is competitive with frequently used much more complex SNV calling algorithms in terms of sensitivity and specificity. It performs specifically well in cases with low allele frequencies. The application to next-generation sequencing data reveals stark differences of the score distributions indicating a strong influence of data specific sources of noise. The proposed model is specifically designed to adjust to these differences.

q-bio.GN↗

arXiv2012

A novel algorithm for simultaneous SNP selection in high-dimensional genome-wide association studies

Background: Identification of causal SNPs in most genome wide association studies relies on approaches that consider each SNP individually. However, there is a strong correlation structure among SNPs that need to be taken into account. Hence, increasingly modern computationally expensive regression methods are employed for SNP selection that consider all markers simultaneously and thus incorporate dependencies among SNPs. Results: We develop a novel multivariate algorithm for large scale SNP selection using CAR score regression, a promising new approach for prioritizing biomarkers. Specifically, we propose a computationally efficient procedure for shrinkage estimation of CAR scores from high-dimensional data. Subsequently, we conduct a comprehensive comparison study including five advanced regression approaches (boosting, lasso, NEG, MCP, and CAR score) and a univariate approach (marginal correlation) to determine the effectiveness in finding true causal SNPs. Conclusions: Simultaneous SNP selection is a challenging task. We demonstrate that our CAR score-based algorithm consistently outperforms all competing approaches, both uni- and multivariate, in terms of correctly recovered causal SNPs and SNP ranking. An R package implementing the approach as well as R code to reproduce the complete study presented here is available from http://strimmerlab.org/software/care/ .

stat.AP↗

arXiv2012

Signal Identification for Rare and Weak Features: Higher Criticism or False Discovery Rates?

Signal identification in large-dimensional settings is a challenging problem in biostatistics. Recently, the method of higher criticism (HC) was shown to be an effective means for determining appropriate decision thresholds. Here, we study HC from a false discovery rate (FDR) perspective. We show that the HC threshold may be viewed as an approximation to a natural class boundary (CB) in two-class discriminant analysis which in turn is expressible as FDR threshold. We demonstrate that in a rare-weak setting in the region of the phase space where signal identification is possible both thresholds are practicably indistinguishable, and thus HC thresholding is identical to using a simple local FDR cutoff. The relationship of the HC and CB thresholds and their properties are investigated both analytically and by simulations, and are further compared by application to four cancer gene expression data sets.

stat.ME↗

arXiv2012

MALDIquant: a versatile R package for the analysis of mass spectrometry data

Summary: MALDIquant is an R package providing a complete and modular analysis pipeline for quantitative analysis of mass spectrometry data. MALDIquant is specifically designed with application in clinical diagnostics in mind and implements sophisticated routines for importing raw data, preprocessing, non-linear peak alignment, and calibration. It also handles technical replicates as well as spectra with unequal resolution. Availability: MALDIquant and its associated R packages readBrukerFlexData and readMzXmlData are freely available from the R archive CRAN (http://cran.r-project.org). The software is distributed under the GNU General Public License (version 3 or later) and is accompanied by example files and data. Additional documentation is available from http://strimmerlab.org/software/maldiquant/.

q-bio.GN↗

arXiv2011

High-dimensional regression and variable selection using CAR scores

Variable selection is a difficult problem that is particularly challenging in the analysis of high-dimensional genomic data. Here, we introduce the CAR score, a novel and highly effective criterion for variable ranking in linear regression based on Mahalanobis-decorrelation of the explanatory variables. The CAR score provides a canonical ordering that encourages grouping of correlated predictors and down-weights antagonistic variables. It decomposes the proportion of variance explained and it is an intermediate between marginal correlation and the standardized regression coefficient. As a population quantity, any preferred inference scheme can be applied for its estimation. Using simulations we demonstrate that variable selection by CAR scores is very effective and yields prediction errors and true and false positive rates that compare favorably with modern regression techniques such as elastic net and boosting. We illustrate our approach by analyzing data concerned with diabetes progression and with the effect of aging on gene expression in the human brain. The R package "care" implementing CAR score regression is available from CRAN.

stat.ME↗

arXiv2011

Introduction to Graphical Modelling

The aim of this chapter is twofold. In the first part we will provide a brief overview of the mathematical and statistical foundations of graphical models, along with their fundamental properties, estimation and basic inference procedures. In particular we will develop Markov networks (also known as Markov random fields) and Bayesian networks, which comprise most past and current literature on graphical models. In the second part we will review some applications of graphical models in systems biology.

stat.ML↗

arXiv2011

Learning false discovery rates by fitting sigmoidal threshold functions

False discovery rates (FDR) are typically estimated from a mixture of a null and an alternative distribution. Here, we study a complementary approach proposed by Rice and Spiegelhalter (2008) that uses as primary quantities the null model and a parametric family for the local false discovery rate. Specifically, we consider the half-normal decay and the beta-uniform mixture models as FDR threshold functions. Using simulations and analysis of real data we compare the performance of the Rice-Spiegelhalter approach with that of competing FDR estimation procedures. If the alternative model is misspecified and an empirical null distribution is employed the accuracy of FDR estimation degrades substantially. Hence, while being a very elegant formalism, the FDR threshold approach requires special care in actual application.

stat.ME↗

arXiv2010

Feature selection in omics prediction problems using cat scores and false nondiscovery rate control

We revisit the problem of feature selection in linear discriminant analysis (LDA), that is, when features are correlated. First, we introduce a pooled centroids formulation of the multiclass LDA predictor function, in which the relative weights of Mahalanobis-transformed predictors are given by correlation-adjusted $t$-scores (cat scores). Second, for feature selection we propose thresholding cat scores by controlling false nondiscovery rates (FNDR). Third, training of the classifier is based on James--Stein shrinkage estimates of correlations and variances, where regularization parameters are chosen analytically without resampling. Overall, this results in an effective and computationally inexpensive framework for high-dimensional prediction with natural feature selection. The proposed shrinkage discriminant procedures are implemented in the R package ``sda'' available from the R repository CRAN.

stat.AP↗

arXiv2009

Entropy inference and the James-Stein estimator, with application to nonlinear gene association networks

We present a procedure for effective estimation of entropy and mutual information from small-sample data, and apply it to the problem of inferring high-dimensional gene association networks. Specifically, we develop a James-Stein-type shrinkage estimator, resulting in a procedure that is highly efficient statistically as well as computationally. Despite its simplicity, we show that it outperforms eight other entropy estimation procedures across a diverse range of sampling scenarios and data-generating models, even in cases of severe undersampling. We illustrate the approach by analyzing E. coli gene expression data and computing an entropy-based gene-association network from gene expression data. A computer program is available that implements the proposed shrinkage estimator.

stat.ML↗

arXiv2009

Gene ranking and biomarker discovery under correlation

Biomarker discovery and gene ranking is a standard task in genomic high throughput analysis. Typically, the ordering of markers is based on a stabilized variant of the t-score, such as the moderated t or the SAM statistic. However, these procedures ignore gene-gene correlations, which may have a profound impact on the gene orderings and on the power of the subsequent tests. We propose a simple procedure that adjusts gene-wise t-statistics to take account of correlations among genes. The resulting correlation-adjusted t-scores ("cat" scores) are derived from a predictive perspective, i.e. as a score for variable selection to discriminate group membership in two-class linear discriminant analysis. In the absence of correlation the cat score reduces to the standard t-score. Moreover, using the cat score it is straightforward to evaluate groups of features (i.e. gene sets). For computation of the cat score from small sample data we propose a shrinkage procedure. In a comparative study comprising six different synthetic and empirical correlation structures we show that the cat score improves estimation of gene orderings and leads to higher power for fixed true discovery rate, and vice versa. Finally, we also illustrate the cat score by analyzing metabolomic data. The shrinkage cat score is implemented in the R package "st" available from URL http://cran.r-project.org/web/packages/st/

stat.AP↗