Principal component analysis based methods in bioinformatics studies

In analysis of bioinformatics data, a unique challenge arises from the high dimensionality of measurements. Without loss of generality, we use genomic study with gene expression measurements as a representative example but note that analysis techniques discussed in this article are also applicable to other types of bioinformatics studies. Principal component analysis (PCA) is a classic dimension reduction approach. It constructs linear combinations of gene expressions, called principal components (PCs). The PCs are orthogonal to each other, can effectively explain variation of gene expressions, and may have a much lower dimensionality. PCA is computationally simple and can be realized using many existing software packages. This article consists of the following parts. First, we review the standard PCA technique and their applications in bioinformatics data analysis. Second, we describe recent 'non-standard' applications of PCA, including accommodating interactions among genes, pathways and network modules and conducting PCA with estimating equations as opposed to gene expressions. Third, we introduce several recently proposed PCA-based techniques, including the supervised PCA, sparse PCA and functional PCA. The supervised PCA and sparse PCA have been shown to have better empirical performance than the standard PCA. The functional PCA can analyze time-course gene expression data. Last, we raise the awareness of several critical but unsolved problems related to PCA. The goal of this article is to make bioinformatics researchers aware of the PCA technique and more importantly its most recent development, so that this simple yet effective dimension reduction technique can be better employed in bioinformatics data analysis.     

Effective dimensionality of large-scale expression data using principal component analysis

Large-scale expression data are today measured for thousands of genes simultaneously. This development is followed by an exploration of theoretical tools to get as much information out of these data as possible. One line is to try to extract the underlying regulatory network. The models used thus far, however, contain many parameters, and a careful investigation is necessary in order not to over-fit the models. We employ principal component analysis to show how, in the context of linear additive models, one can get a rough estimate of the effective dimensionality (the number of information-carrying dimensions) of large-scale gene expression datasets. We treat both the lack of independence of different measurements in a time series and the fact that that measurements are subject to some level of noise, both of which reduce the effective dimensionality and thereby constrain the complexity of models which can be built from the data.     

Statistical analysis of a small set of time-ordered gene expression data using linear splines

MOTIVATION: Recently, the temporal response of genes to changes in their environment has been investigated using cDNA microarray technology by measuring the gene expression levels at a small number of time points. Conventional techniques for time series analysis are not suitable for such a short series of time-ordered data. The analysis of gene expression data has therefore usually been limited to a fold-change analysis, instead of a systematic statistical approach. METHODS: We use the maximum likelihood method together with Akaike's Information Criterion to fit linear splines to a small set of time-ordered gene expression data in order to infer statistically meaningful information from the measurements. The significance of measured gene expression data is assessed using Student's t-test. RESULTS: Previous gene expression measurements of the cyanobacterium Synechocystis sp. PCC6803 were reanalyzed using linear splines. The temporal response was identified of many genes that had been missed by a fold-change analysis. Based on our statistical analysis, we found that about four gene expression measurements or more are needed at each time point.     

Interval-valued analysis for discriminative gene selection and tissue sample classification using microarray data

An important application of gene expression data is to classify samples in a variety of diagnostic fields. However, high dimensionality and a small number of noisy samples pose significant challenges to existing classification methods. Focused on the problems of overfitting and sensitivity to noise of the dataset in the classification of microarray data, we propose an interval-valued analysis method based on a rough set technique to select discriminative genes and to use these genes to classify tissue samples of microarray data. We first select a small subset of genes based on interval-valued rough set by considering the preference-ordered domains of the gene expression data, and then classify test samples into certain classes with a term of similar degree. Experiments show that the proposed method is able to reach high prediction accuracies with a small number of selected genes and its performance is robust to noise.     

Supervised principal component analysis for gene set enrichment of microarray data with continuous or survival outcomes

MOTIVATION: Gene set analysis allows formal testing of subtle but coordinated changes in a group of genes, such as those defined by Gene Ontology (GO) or KEGG Pathway databases. We propose a new method for gene set analysis that is based on principal component analysis (PCA) of genes expression values in the gene set. PCA is an effective method for reducing high dimensionality and capture variations in gene expression values. However, one limitation with PCA is that the latent variable identified by the first PC may be unrelated to outcome. RESULTS: In the proposed supervised PCA (SPCA) model for gene set analysis, the PCs are estimated from a selected subset of genes that are associated with outcome. As outcome information is used in the gene selection step, this method is supervised, thus called the Supervised PCA model. Because of the gene selection step, test statistic in SPCA model can no longer be approximated well using t-distribution. We propose a two-component mixture distribution based on Gumbel exteme value distributions to account for the gene selection step. We show the proposed method compares favorably to currently available gene set analysis methods using simulated and real microarray data. SOFTWARE: The R code for the analysis used in this article are available upon request, we are currently working on implementing the proposed method in an R package.     

Accounting for probe-level noise in principal component analysis of microarray data

MOTIVATION: Principal Component Analysis (PCA) is one of the most popular dimensionality reduction techniques for the analysis of high-dimensional datasets. However, in its standard form, it does not take into account any error measures associated with the data points beyond a standard spherical noise. This indiscriminate nature provides one of its main weaknesses when applied to biological data with inherently large variability, such as expression levels measured with microarrays. Methods now exist for extracting credibility intervals from the probe-level analysis of cDNA and oligonucleotide microarray experiments. These credibility intervals are gene and experiment specific, and can be propagated through an appropriate probabilistic downstream analysis. RESULTS: We propose a new model-based approach to PCA that takes into account the variances associated with each gene in each experiment. We develop an efficient EM-algorithm to estimate the parameters of our new model. The model provides significantly better results than standard PCA, while remaining computationally reasonable. We show how the model can be used to 'denoise' a microarray dataset leading to improved expression profiles and tighter clustering across profiles. The probabilistic nature of the model means that the correct number of principal components is automatically obtained.     

Time-course analysis of cyanobacterium transcriptome: detecting oscillatory genes

The microarray technique allows the simultaneous measurements of the expression levels of thousands of mRNAs. By mining these data one can identify the dynamics of the gene expression time series. The detection of genes that are periodically expressed is an important step that allows us to study the regulatory mechanisms associated with the circadian cycle. The problem of finding periodicity in biological time series poses many challenges. Such challenge occurs due to the fact that the observed time series usually exhibit non-idealities, such as noise, short length, outliers and unevenly sampled time points. Consequently, the method for finding periodicity should preferably be robust against such anomalies in the data. In this paper, we propose a general and robust procedure for identifying genes with a periodic signature at a given significance level. This identification method is based on autoregressive models and the information theory. By using simulated data we show that the suggested method is capable of identifying rhythmic profiles even in the presence of noise and when the number of data points is small. By recourse of our analysis, we uncover the circadian rhythmic patterns underlying the gene expression profiles from Cyanobacterium Synechocystis.     

Using interpolation to estimate system uncertainty in gene expression experiments

The widespread use of high-throughput experimental assays designed to measure the entire complement of a cell's genes or gene products has led to vast stores of data that are extremely plentiful in terms of the number of items they can measure in a single sample, yet often sparse in the number of samples per experiment due to their high cost. This often leads to datasets where the number of treatment levels or time points sampled is limited, or where there are very small numbers of technical and/or biological replicates. Here we introduce a novel algorithm to quantify the uncertainty in the unmeasured intervals between biological measurements taken across a set of quantitative treatments. The algorithm provides a probabilistic distribution of possible gene expression values within unmeasured intervals, based on a plausible biological constraint. We show how quantification of this uncertainty can be used to guide researchers in further data collection by identifying which samples would likely add the most information to the system under study. Although the context for developing the algorithm was gene expression measurements taken over a time series, the approach can be readily applied to any set of quantitative systems biology measurements taken following quantitative (i.e. non-categorical) treatments. In principle, the method could also be applied to combinations of treatments, in which case it could greatly simplify the task of exploring the large combinatorial space of future possible measurements.     

Feature Selection for high Dimensional DNA Microarray data using hybrid approaches

Feature selection from DNA microarray data is a major challenge due to high dimensionality in expression data. The number of samples in the microarray data set is much smaller compared to the number of genes. Hence the data is improper to be used as the training set of a classifier. Therefore it is important to select features prior to training the classifier. It should be noted that only a small subset of genes from the data set exhibits a strong correlation with the class. This is because finding the relevant genes from the data set is often non-trivial. Thus there is a need to develop robust yet reliable methods for gene finding in expression data. We describe the use of several hybrid feature selection approaches for gene finding in expression data. These approaches include filtering (filter out the best genes from the data set) and wrapper (best subset of genes from the data set) phases. The methods use information gain (IG) and Pearson Product Moment Correlation (PPMC) as the filtering parameters and biogeography based optimization (BBO) as the wrapper approach. K nearest neighbour algorithm (KNN) and back propagation neural network are used for evaluating the fitness of gene subsets during feature selection. Our analysis shows that an impressive performance is provided by the IG-BBO-KNN combination in different data sets with high accuracy (>90%) and low error rate.     

A weighted principal component analysis and its application to gene expression data

In this work, we introduce in the first part new developments in Principal Component Analysis (PCA) and in the second part a new method to select variables (genes in our application). Our focus is on problems where the values taken by each variable do not all have the same importance and where the data may be contaminated with noise and contain outliers, as is the case with microarray data. The usual PCA is not appropriate to deal with this kind of problems. In this context, we propose the use of a new correlation coefficient as an alternative to Pearson's. This leads to a so-called weighted PCA (WPCA). In order to illustrate the features of our WPCA and compare it with the usual PCA, we consider the problem of analyzing gene expression data sets. In the second part of this work, we propose a new PCA-based algorithm to iteratively select the most important genes in a microarray data set. We show that this algorithm produces better results when our WPCA is used instead of the usual PCA. Furthermore, by using Support Vector Machines, we show that it can compete with the Significance Analysis of Microarrays algorithm.     

Systematic benchmarking of microarray data classification: assessing the role of non-linearity and dimensionality reduction

MOTIVATION: Microarrays are capable of determining the expression levels of thousands of genes simultaneously. In combination with classification methods, this technology can be useful to support clinical management decisions for individual patients, e.g. in oncology. The aim of this paper is to systematically benchmark the role of non-linear versus linear techniques and dimensionality reduction methods. RESULTS: A systematic benchmarking study is performed by comparing linear versions of standard classification and dimensionality reduction techniques with their non-linear versions based on non-linear kernel functions with a radial basis function (RBF) kernel. A total of 9 binary cancer classification problems, derived from 7 publicly available microarray datasets, and 20 randomizations of each problem are examined. CONCLUSIONS: Three main conclusions can be formulated based on the performances on independent test sets. (1) When performing classification with least squares support vector machines (LS-SVMs) (without dimensionality reduction), RBF kernels can be used without risking too much overfitting. The results obtained with well-tuned RBF kernels are never worse and sometimes even statistically significantly better compared to results obtained with a linear kernel in terms of test set receiver operating characteristic and test set accuracy performances. (2) Even for classification with linear classifiers like LS-SVM with linear kernel, using regularization is very important. (3) When performing kernel principal component analysis (kernel PCA) before classification, using an RBF kernel for kernel PCA tends to result in overfitting, especially when using supervised feature selection. It has been observed that an optimal selection of a large number of features is often an indication for overfitting. Kernel PCA with linear kernel gives better results.     

Identifying the Genetic Variation of Gene Expression Using Gene Sets: Application of Novel Gene Set eQTL Approach to PharmGKB and KEGG

Genetic variation underlying the regulation of mRNA gene expression in humans may provide key insights into the molecular mechanisms of human traits and complex diseases. Current statistical methods to map genetic variation associated with mRNA gene expression have typically applied standard linkage and/or association methods; however, when genome-wide SNP and mRNA expression data are available performing all pair wise comparisons is computationally burdensome and may not provide optimal power to detect associations. Consideration of different approaches to account for the high dimensionality and multiple testing issues may provide increased efficiency and statistical power. Here we present a novel approach to model and test the association between genetic variation and mRNA gene expression levels in the context of gene sets (GSs) and pathways, referred to as gene set - expression quantitative trait loci analysis (GS-eQTL). The method uses GSs to initially group SNPs and mRNA expression, followed by the application of principal components analysis (PCA) to collapse the variation and reduce the dimensionality within the GSs. We applied GS-eQTL to assess the association between SNP and mRNA expression level data collected from a cell-based model system using PharmGKB and KEGG defined GSs. We observed a large number of significant GS-eQTL associations, in which the most significant associations arose between genetic variation and mRNA expression from the same GS. However, a number of associations involving genetic variation and mRNA expression from different GSs were also identified. Our proposed GS-eQTL method effectively addresses the multiple testing limitations in eQTL studies and provides biological context for SNP-expression associations.     

Integrating heterogeneous gene expression data for gene regulatory network modelling

Gene regulatory networks (GRNs) are complex biological systems that have a large impact on protein levels, so that discovering network interactions is a major objective of systems biology. Quantitative GRN models have been inferred, to date, from time series measurements of gene expression, but at small scale, and with limited application to real data. Time series experiments are typically short (number of time points of the order of ten), whereas regulatory networks can be very large (containing hundreds of genes). This creates an under-determination problem, which negatively influences the results of any inferential algorithm. Presented here is an integrative approach to model inference, which has not been previously discussed to the authors' knowledge. Multiple heterogeneous expression time series are used to infer the same model, and results are shown to be more robust to noise and parameter perturbation. Additionally, a wavelet analysis shows that these models display limited noise over-fitting within the individual datasets.     

A blocking strategy to improve gene selection for classification of gene expression data

Because of high dimensionality, machine learning algorithms typically rely on feature selection techniques in order to perform effective classification in microarray gene expression data sets. However, the large number of features compared to the number of samples makes the task of feature selection computationally hard and prone to errors. This paper interprets feature selection as a task of stochastic optimization, where the goal is to select among an exponential number of alternative gene subsets the one expected to return the highest generalization in classification. Blocking is an experimental design strategy which produces similar experimental conditions to compare alternative stochastic configurations in order to be confident that observed differences in accuracy are due to actual differences rather than to fluctuations and noise effects. We propose an original blocking strategy for improving feature selection which aggregates in a paired way the validation outcomes of several learning algorithms to assess a gene subset and compare it to others. This is a novelty with respect to conventional wrappers, which commonly adopt a sole learning algorithm to evaluate the relevance of a given set of variables. The rationale of the approach is that, by increasing the amount of experimental conditions under which we validate a feature subset, we can lessen the problems related to the scarcity of samples and consequently come up with a better selection. The paper shows that the blocking strategy significantly improves the performance of a conventional forward selection for a set of 16 publicly available cancer expression data sets. The experiments involve six different classifiers and show that improvements take place independent of the classification algorithm used after the selection step. Two further validations based on available biological annotation support the claim that blocking strategies in feature selection may improve the accuracy and the quality of the solution. The first validation is based on retrieving PubMEd abstracts associated to the selected genes and matching them to regular expressions describing the biological phenomenon underlying the expression data sets. The biological validation that follows is based on the use of the Bioconductor package GoStats in order to perform Gene Ontology statistical analysis.     

Dynamic models of gene expression and classification

Powerful new methods, like expression profiles using cDNA arrays, have been used to monitor changes in gene expression levels as a result of a variety of metabolic, xenobiotic or pathogenic challenges. This potentially vast quantity of data enables, in principle, the dissection of the complex genetic networks that control the patterns and rhythms of gene expression in the cell. Here we present a general approach to developing dynamic models for analyzing time series of whole genome expression. In this approach, a self-consistent calculation is performed that involves both linear and non-linear response terms for interrelating gene expression levels. This calculation uses singular value decomposition (SVD) not as a statistical tool but as a means of inverting noisy and near-singular matrices. The linear transition matrix that is determined from this calculation can be used to calculate the underlying network reflected in the data. This suggests a direct method of classifying genes according to their place in the resulting network. In addition to providing a means to model such a large multivariate system this approach can be used to reduce the dimensionality of the problem in a rational and consistent way, and suppress the strong noise amplification effects often encountered with expression profile data. Non-linear and higher-order Markov behavior of the network are also determined in this self-consistent method. In data sets from yeast, we calculate the Markov matrix and the gene classes based on the linear-Markov network. These results compare favorably with previously used methods like cluster analysis. Our dynamic method appears to give a broad and general framework for data analysis and modeling of gene expression arrays. Electronic Publication     

Reconstructing the temporal ordering of biological samples using microarray data

MOTIVATION: Accurate time series for biological processes are difficult to estimate due to problems of synchronization, temporal sampling and rate heterogeneity. Methods are needed that can utilize multi-dimensional data, such as those resulting from DNA microarray experiments, in order to reconstruct time series from unordered or poorly ordered sets of observations. RESULTS: We present a set of algorithms for estimating temporal orderings from unordered sets of sample elements. The techniques we describe are based on modifications of a minimum-spanning tree calculated from a weighted, undirected graph. We demonstrate the efficacy of our approach by applying these techniques to an artificial data set as well as several gene expression data sets derived from DNA microarray experiments. In addition to estimating orderings, the techniques we describe also provide useful heuristics for assessing relevant properties of sample datasets such as noise and sampling intensity, and we show how a data structure called a PQ-tree can be used to represent uncertainty in a reconstructed ordering. AVAILABILITY: Academic implementations of the ordering algorithms are available as source code (in the programming language Python) on our web site, along with documentation on their use. The artificial 'jelly roll' data set upon which the algorithm was tested is also available from this web site. The publicly available gene expression data may be found at http://genome-www.stanford.edu/cellcycle/ and http://caulobacter.stanford.edu/CellCycle/.     

Integrative analysis of gene expression and copy number alterations using canonical correlation analysis

Background  

With the rapid development of new genetic measurement methods, several types of genetic alterations can be quantified in a high-throughput manner. While the initial focus has been on investigating each data set separately, there is an increasing interest in studying the correlation structure between two or more data sets. Multivariate methods based on Canonical Correlation Analysis (CCA) have been proposed for integrating paired genetic data sets. The high dimensionality of microarray data imposes computational difficulties, which have been addressed for instance by studying the covariance structure of the data, or by reducing the number of variables prior to applying the CCA. In this work, we propose a new method for analyzing high-dimensional paired genetic data sets, which mainly emphasizes the correlation structure and still permits efficient application to very large data sets. The method is implemented by translating a regularized CCA to its dual form, where the computational complexity depends mainly on the number of samples instead of the number of variables. The optimal regularization parameters are chosen by cross-validation. We apply the regularized dual CCA, as well as a classical CCA preceded by a dimension-reducing Principal Components Analysis (PCA), to a paired data set of gene expression changes and copy number alterations in leukemia.     

Harmony: EEG/MEG Linear Inverse Source Reconstruction in the Anatomical Basis of Spherical Harmonics

EEG/MEG source localization based on a “distributed solution” is severely underdetermined, because the number of sources is much larger than the number of measurements. In particular, this makes the solution strongly affected by sensor noise. A new way to constrain the problem is presented. By using the anatomical basis of spherical harmonics (or spherical splines) instead of single dipoles the dimensionality of the inverse solution is greatly reduced without sacrificing the quality of the data fit. The smoothness of the resulting solution reduces the surface bias and scatter of the sources (incoherency) compared to the popular minimum-norm algorithms where single-dipole basis is used (MNE, depth-weighted MNE, dSPM, sLORETA, LORETA, IBF) and allows to efficiently reduce the effect of sensor noise. This approach, termed Harmony, performed well when applied to experimental data (two exemplars of early evoked potentials) and showed better localization precision and solution coherence than the other tested algorithms when applied to realistically simulated data.