Analysis Of High-throughput Genomic Data Research Articles

PSF toolkit: an R package for pathway curation and topology-aware analysis.

Most high throughput genomic data analysis pipelines currently rely on over-representation or gene set enrichment analysis (ORA/GSEA) approaches for functional analysis. In contrast, topology-based pathway analysis methods, which offer a more biologically informed perspective by incorporating interaction and topology information, have remained underutilized and inaccessible due to various limiting factors. These methods heavily rely on the quality of pathway topologies and often utilize predefined topologies from databases without assessing their correctness. To address these issues and make topology-aware pathway analysis more accessible and flexible, we introduce the PSF (Pathway Signal Flow) toolkit R package. Our toolkit integrates pathway curation and topology-based analysis, providing interactive and command-line tools that facilitate pathway importation, correction, and modification from diverse sources. This enables users to perform topology-based pathway signal flow analysis in both interactive and command-line modes. To showcase the toolkit's usability, we curated 36 KEGG signaling pathways and conducted several use-case studies, comparing our method with ORA and the topology-based signaling pathway impact analysis (SPIA) method. The results demonstrate that the algorithm can effectively identify ORA enriched pathways while providing more detailed branch-level information. Moreover, in contrast to the SPIA method, it offers the advantage of being cut-off free and less susceptible to the variability caused by selection thresholds. By combining pathway curation and topology-based analysis, the PSF toolkit enhances the quality, flexibility, and accessibility of topology-aware pathway analysis. Researchers can now easily import pathways from various sources, correct and modify them as needed, and perform detailed topology-based pathway signal flow analysis. In summary, our PSF toolkit offers an integrated solution that addresses the limitations of current topology-based pathway analysis methods. By providing interactive and command-line tools for pathway curation and topology-based analysis, we empower researchers to conduct comprehensive pathway analyses across a wide range of applications.

Accurate and fast small p-value estimation for permutation tests in high-throughput genomic data analysis with the cross-entropy method.

Permutation tests are widely used for statistical hypothesis testing when the sampling distribution of the test statistic under the null hypothesis is analytically intractable or unreliable due to finite sample sizes. One critical challenge in the application of permutation tests in genomic studies is that an enormous number of permutations are often needed to obtain reliable estimates of very small p-values, leading to intensive computational effort. To address this issue, we develop algorithms for the accurate and efficient estimation of small p-values in permutation tests for paired and independent two-group genomic data, and our approaches leverage a novel framework for parameterizing the permutation sample spaces of those two types of data respectively using the Bernoulli and conditional Bernoulli distributions, combined with the cross-entropy method. The performance of our proposed algorithms is demonstrated through the application to two simulated datasets and two real-world gene expression datasets generated by microarray and RNA-Seq technologies and comparisons to existing methods such as crude permutations and SAMC, and the results show that our approaches can achieve orders of magnitude of computational efficiency gains in estimating small p-values. Our approaches offer promising solutions for the improvement of computational efficiencies of existing permutation test procedures and the development of new testing methods using permutations in genomic data analysis.

L0-Regularized Learning for High-Dimensional Additive Hazards Regression

Sparse learning in high-dimensional survival analysis is of great practical importance, as exemplified by modern applications in credit risk analysis and high-throughput genomic data analysis. In this article, we consider the L0-regularized learning for simultaneous variable selection and estimation under the framework of additive hazards models and utilize the idea of primal dual active sets to develop an algorithm targeted at solving the traditionally nonpolynomial time optimization problem. Under interpretable conditions, comprehensive statistical properties, including model selection consistency, oracle inequalities under various estimation losses, and the oracle property, are established for the global optimizer of the proposed approach. Moreover, our theoretical analysis for the algorithmic solution reveals that the proposed L0-regularized learning can be more efficient than other regularization methods in that it requests a smaller sample size as well as a lower minimum signal strength to identify the significant features. The effectiveness of the proposed method is evidenced by simulation studies and real-data analysis. Summary of Contribution: Feature selection is a fundamental statistical learning technique under high dimensions and routinely encountered in various areas, including operations research and computing. This paper focuses on the L0-regularized learning for feature selection in high-dimensional additive hazards regression. The matching algorithm for solving the nonconvex L0-constrained problem is scalable and enjoys comprehensive theoretical properties.

Comparative Study of three Clustering Algorithms for Microarray Data

High throughput genomic data analysis is becoming an increasingly integral part of biomedical research. The information derived from gene expression analysis helps in diagnosing the treatment modality given to the patient. However, the amount of data is humongous and becomes complex to examine manually. Unsupervised machine learning algorithms perform complex tasks on an unlabelled data by clustering to comprehend the underlying structure and behaviour of the pattern. Clustering microarray data, examines the differential expressed genes found by grouping the genes based on the similarity of the expression values. In this study, we propose to elucidate the best clustering algorithm for gene expression data on various clinical conditions. The proposed study was carried on three gene expression datasets of Severe acute respiratory syndrome, Amyotrophic lateral sclerosis and Parkinson’s disease. Differentially expressed genes were found at three p-values 0.01, 0.05, 0.001 and the most significant number of genes were retrieved at p-value 0.05. We experimented the differential expressed genes on three clustering algorithms, namely Hierarchical clustering, k-means clustering and fuzzy clustering of the three diseases. The performance of the three clustering algorithms was evaluated using the internal validity index, wherein Hierarchical clustering was found to be best for gene expression data.

ExperimentSubset: an R package to manage subsets of Bioconductor Experiment objects.

R Experiment objects such as the SummarizedExperiment or SingleCellExperiment are data containers for storing one or more matrix-like assays along with associated row and column data. These objects have been used to facilitate the storage and analysis of high-throughput genomic data generated from technologies such as single-cell RNA sequencing. One common computational task in many genomics analysis workflows is to perform subsetting of the data matrix before applying down-stream analytical methods. For example, one may need to subset the columns of the assay matrix to exclude poor-quality samples or subset the rows of the matrix to select the most variable features. Traditionally, a second object is created that contains the desired subset of assay from the original object. However, this approach is inefficient as it requires the creation of an additional object containing a copy of the original assay and leads to challenges with data provenance. To overcome these challenges, we developed an R package called ExperimentSubset, which is a data container that implements classes for efficient storage and streamlined retrieval of assays that have been subsetted by rows and/or columns. These classes are able to inherently provide data provenance by maintaining the relationship between the subsetted and parent assays. We demonstrate the utility of this package on a single-cell RNA-seq dataset by storing and retrieving subsets at different stages of the analysis while maintaining a lower memory footprint. Overall, the ExperimentSubset is a flexible container for the efficient management of subsets. ExperimentSubset package is available at Bioconductor: https://bioconductor.org/packages/ExperimentSubset/ and Github: https://github.com/campbio/ExperimentSubset. Supplementary data are available at Bioinformatics online.

First person – Phillipe O'Brien, Kai Guo, Stephanie Eid, Amy Rumora and Lucy Hinder

ABSTRACTFirst Person is a series of interviews with the first authors of a selection of papers published in Disease Models & Mechanisms (DMM), helping early-career researchers promote themselves alongside their papers. Phillipe O'Brien, Kai Guo, Stephanie Eid, Amy Rumora and Lucy Hinder are first authors on ‘Integrated lipidomic and transcriptomic analyses identify altered nerve triglycerides in mouse models of prediabetes and type 2 diabetes’, published in DMM. Phillipe, Stephanie, Amy and Lucy are all postdoctoral researchers in the lab of Eva L. Feldman at the University of Michigan, MI, USA, investigating the pathogenesis of peripheral neuropathy using mouse models of obesity and diabetes. Kai is a postdoctoral researcher in the lab of Junguk Hur at the University of North Dakota, Grand Forks, ND, USA, and uses high-throughput genomic data analysis to understand the development of neurological disorders.

Learning an un-supervised – Clustering algorithm Monte Carlo over Consensus Clustering for Genomic Data for Tumor Identification

Clustering involves the grouping of similar objects into a set known as cluster. Objects in one cluster are likely to be different when compared to objects grouped under another cluster. Gene expression is the process by which information from a gene is used in the synthesis of a functional gene product. Subgroup classification is a basic task in high-throughput genomic data analysis, especially for gene expression and methylation data analysis. Mostly, unsupervised clustering methods are applied to predict new subgroups or test the consistency with known annotations. To get a stable classification of subgroups, consensus clustering is always performed. It clusters repeatedly with a randomly sampled subset of data and summarizes the robustness of the clustering. When faced with significant uncertainty in the process of making a forecast or estimation, the Monte Carlo Simulation might prove to be a better solution. Monte Carlo3C is a consensus clustering algorithm that uses a Monte Carlo simulation to eliminate overfitting and can reject the null hypothesis when only one cluster is there.

Gpart: human genome partitioning and visualization of high-density SNP data by identifying haplotype blocks.

SummaryFor the analysis of high-throughput genomic data produced by next-generation sequencing (NGS) technologies, researchers need to identify linkage disequilibrium (LD) structure in the genome. In this work, we developed an R package gpart which provides clustering algorithms to define LD blocks or analysis units consisting of SNPs. The visualization tool in gpart can display the LD structure and gene positions for up to 20 000 SNPs in one image. The gpart functions facilitate construction of LD blocks and SNP partitions for vast amounts of genome sequencing data within reasonable time and memory limits in personal computing environments.Availability and implementationThe R package is available at https://bioconductor.org/packages/gpart.Supplementary information Supplementary data are available at Bioinformatics online.

Penalized empirical likelihood inference for sparse additive hazards regression with a diverging number of covariates

High-dimensional sparse modeling with censored survival data is of great practical importance, as exemplified by applications in high-throughput genomic data analysis. In this paper, we propose a class of regularization methods, integrating both the penalized empirical likelihood and pseudoscore approaches, for variable selection and estimation in sparse and high-dimensional additive hazards regression models. When the number of covariates grows with the sample size, we establish asymptotic properties of the resulting estimator and the oracle property of the proposed method. It is shown that the proposed estimator is more efficient than that obtained from the non-concave penalized likelihood approach in the literature. Based on a penalized empirical likelihood ratio statistic, we further develop a nonparametric likelihood approach for testing the linear hypothesis of regression coefficients and constructing confidence regions consequently. Simulation studies are carried out to evaluate the performance of the proposed methodology and also two real data sets are analyzed.

Practical impacts of genomic data "cleaning" on biological discovery using surrogate variable analysis.

BackgroundGenomic data production is at its highest level and continues to increase, making available novel primary data and existing public data to researchers for exploration. Here we explore the consequences of “batch” correction for biological discovery in two publicly available expression datasets. We consider this to include the estimation of and adjustment for wide-spread systematic heterogeneity in genomic measurements that is unrelated to the effects under study, whether it be technical or biological in nature.MethodsWe present three illustrative data analyses using surrogate variable analysis (SVA) and describe how to perform artifact discovery in light of natural heterogeneity within biological groups, secondary biological questions of interest, and non-linear treatment effects in a dataset profiling differentiating pluripotent cells (GSE32923) and another from human brain tissue (GSE30272).ResultsCareful specification of biological effects of interest is very important to factor-based approaches like SVA. We demonstrate greatly sharpened global and gene-specific differential expression across treatment groups in stem cell systems. Similarly, we demonstrate how to preserve major non-linear effects of age across the lifespan in the brain dataset. However, the gains in precisely defining known effects of interest come at the cost of much other information in the “cleaned” data, including sex, common copy number effects and sample or cell line-specific molecular behavior.ConclusionsOur analyses indicate that data “cleaning” can be an important component of high-throughput genomic data analysis when interrogating explicitly defined effects in the context of data affected by robust technical artifacts. However, caution should be exercised to avoid removing biological signal of interest. It is also important to note that open data exploration is not possible after such supervised “cleaning”, because effects beyond those stipulated by the researcher may have been removed. With the goal of making these statistical algorithms more powerful and transparent to researchers in the biological sciences, we provide exploratory plots and accompanying R code for identifying and guiding “cleaning” process (https://github.com/andrewejaffe/StemCellSVA). The impact of these methods is significant enough that we have made newly processed data available for the brain data set at http://braincloud.jhmi.edu/plots/ and GSE30272.Electronic supplementary materialThe online version of this article (doi:10.1186/s12859-015-0808-5) contains supplementary material, which is available to authorized users.

Glmgraph: an R package for variable selection and predictive modeling of structured genomic data.

One central theme of modern high-throughput genomic data analysis is to identify relevant genomic features as well as build up a predictive model based on selected features for various tasks such as personalized medicine. Correlating the large number of 'omics' features with a certain phenotype is particularly challenging due to small sample size (n) and high dimensionality (p). To address this small n, large p problem, various forms of sparse regression models have been proposed by exploiting the sparsity assumption. Among these, network-constrained sparse regression model is of particular interest due to its ability to utilize the prior graph/network structure in the omics data. Despite its potential usefulness for omics data analysis, no efficient R implementation is publicly available. Here we present an R software package 'glmgraph' that implements the graph-constrained regularization for both sparse linear regression and sparse logistic regression. We implement both the L1 penalty and minimax concave penalty for variable selection and Laplacian penalty for coefficient smoothing. Efficient coordinate descent algorithm is used to solve the optimization problem. We demonstrate the use of the package by applying it to a human microbiome dataset, where phylogeny structure among bacterial taxa is available. 'glmgraph' is implemented in R and C++ Armadillo and publicly available under CRAN.

DupChecker: a bioconductor package for checking high-throughput genomic data redundancy in meta-analysis.

BackgroundMeta-analysis has become a popular approach for high-throughput genomic data analysis because it often can significantly increase power to detect biological signals or patterns in datasets. However, when using public-available databases for meta-analysis, duplication of samples is an often encountered problem, especially for gene expression data. Not removing duplicates could lead false positive finding, misleading clustering pattern or model over-fitting issue, etc in the subsequent data analysis.ResultsWe developed a Bioconductor package Dupchecker that efficiently identifies duplicated samples by generating MD5 fingerprints for raw data. A real data example was demonstrated to show the usage and output of the package.ConclusionsResearchers may not pay enough attention to checking and removing duplicated samples, and then data contamination could make the results or conclusions from meta-analysis questionable. We suggest applying DupChecker to examine all gene expression data sets before any data analysis step.Electronic supplementary materialThe online version of this article (doi:10.1186/1471-2105-15-323) contains supplementary material, which is available to authorized users.

Unsupervised outlier profile analysis.

In much of the analysis of high-throughput genomic data, “interesting” genes have been selected based on assessment of differential expression between two groups or generalizations thereof. Most of the literature focuses on changes in mean expression or the entire distribution. In this article, we explore the use of C(α) tests, which have been applied in other genomic data settings. Their use for the outlier expression problem, in particular with continuous data, is problematic but nevertheless motivates new statistics that give an unsupervised analog to previously developed outlier profile analysis approaches. Some simulation studies are used to evaluate the proposal. A bivariate extension is described that can accommodate data from two platforms on matched samples. The proposed methods are applied to data from a prostate cancer study.

Exploiting parallel R in the cloud with SPRINT.

Advances in DNA Microarray devices and next-generation massively parallel DNA sequencing platforms have led to an exponential growth in data availability but the arising opportunities require adequate computing resources. High Performance Computing (HPC) in the Cloud offers an affordable way of meeting this need. Bioconductor, a popular tool for high-throughput genomic data analysis, is distributed as add-on modules for the R statistical programming language but R has no native capabilities for exploiting multi-processor architectures. SPRINT is an R package that enables easy access to HPC for genomics researchers. This paper investigates: setting up and running SPRINT-enabled genomic analyses on Amazon's Elastic Compute Cloud (EC2), the advantages of submitting applications to EC2 from different parts of the world and, if resource underutilization can improve application performance. The SPRINT parallel implementations of correlation, permutation testing, partitioning around medoids and the multi-purpose papply have been benchmarked on data sets of various size on Amazon EC2. Jobs have been submitted from both the UK and Thailand to investigate monetary differences. It is possible to obtain good, scalable performance but the level of improvement is dependent upon the nature of the algorithm. Resource underutilization can further improve the time to result. End-user's location impacts on costs due to factors such as local taxation. Although not designed to satisfy HPC requirements, Amazon EC2 and cloud computing in general provides an interesting alternative and provides new possibilities for smaller organisations with limited funds.

High-Dimensional Sparse Additive Hazards Regression

High-dimensional sparse modeling with censored survival data is of great practical importance, as exemplified by modern applications in high-throughput genomic data analysis and credit risk analysis. In this article, we propose a class of regularization methods for simultaneous variable selection and estimation in the additive hazards model, by combining the nonconcave penalized likelihood approach and the pseudoscore method. In a high-dimensional setting where the dimensionality can grow fast, polynomially or nonpolynomially, with the sample size, we establish the weak oracle property and oracle property under mild, interpretable conditions, thus providing strong performance guarantees for the proposed methodology. Moreover, we show that the regularity conditions required by the L 1 method are substantially relaxed by a certain class of sparsity-inducing concave penalties. As a result, concave penalties such as the smoothly clipped absolute deviation, minimax concave penalty, and smooth integration of counting and absolute deviation can significantly improve on the L 1 method and yield sparser models with better prediction performance. We present a coordinate descent algorithm for efficient implementation and rigorously investigate its convergence properties. The practical use and effectiveness of the proposed methods are demonstrated by simulation studies and a real data example.

A bioinformatics framework for the identification of active regulatory elements through the integrative analysis of high-throughput genomic data

Motivations. High-throughput technologies as microarray, Cap Analysis of Gene Expression (CAGE), and chromatin immunoprecipitation (ChIP), coupled with next generation sequencing (NGS) allow the identification of the molecular mechanisms regulating DNA transcription. In particular, CAGE and chromatin immunoprecipitation followed by DNA sequencing (ChIP-seq) are particularly suited to analyze transcriptional regulation and DNA methylation and histone modification patterns, while gene expression microarrays allow assessing transcriptional patterns. All these technologies produce data that are highly informative per se, but merging and integrating the various types of information would help answering many long-standing questions related to fundamental mechanisms of gene regulation and genome utilization. Data integration can be addressed using three main approaches, i.e., reduction of data complexity, unsupervised integration, and supervised integration [1]. Although these methods proved their efficiency in the analysis of single types of data, no bioinformatics tool integrates them all in a unified pipeline. In this context, we present a bioinformatics framework for the integrative analysis of CAGE, ChIP-Seq, and microarray data and the genome-wide identification of active regulatory elements. The computational workflow comprises three steps, i.e., i) stand-alone analysis of CAGE, ChIP-Seq, and microarray data; ii) construction of a unified data structure where results from CAGE, ChIP-Seq, and microarray analysis converge to generate integrated patterns of genomic signals; and iii) analysis of the integrated patterns to identify active regulatory elements. Methods. CAGE data have been analyzed using standard methods [2]. ChIP-seq peak calling was obtained using SICER [3]. Microarray data have been processed using PREDA for the detection of chromosomal patterns of gene expression [4]. Results from stand-alone analyses have been integrated in a unique matrix based on the UCSC BED file structure and on the chromosomal coordinate of genomic elements. The analysis of the integrated matrix has been performed using an ad-hoc procedure coded in R for the identification of i) enhancers, ii) non-coding RNA, and iii) promoters. The algorithm comprises three steps: in the first, CAGE and ChIP-seq peaks are merged to determine which histone modifications are present in correspondence of a CAGE peak. CAGE peaks that fall within H3K4me1 regions are classified as associated with a putative enhancer, while peaks that fall into both H3K4me1 and H3K4me3 regions are classified as associated with a putative promoter. In the second step, the algorithm verifies if there exists bi-directional transcription between two consecutive CAGE peaks located on opposite strands. Briefly, given any two peaks on opposite strands, an R function calculates the distance between them and labels as bi-directionally transcribed those peaks lying at a distance ≤ 1 kb. All other peaks are marked as mono-directionally transcribed. In the last step, the presence of putative enhancers and promoters is linked to the local pattern of gene expression. Results. The pipeline has been applied for the genome-wide characterization of mono-directionally and bi-directionally transcribed promoters and enhancers involved in self-renewal, commitment, and differentiation of human stem/progenitor cells and their progeny. Specifically, we analyzed data from cord-blood derived hematopoietic stem cells (HSCs) and lineage-restricted erythroblasts and myelomonocytes identifying specific regulatory regions differentially engaged during HSC differentiation. This work is supported by FIRB – Programma “Futuro in Ricerca” 2010 (RBFR10OS4G) and Progetto Bandiera Epigenomica.

Using offspring-parent triads to study complex traits: A tutorial based on orofacial clefts

Health investigators routinely collect DNA and environmental data from study participants in order to assess the impact of genetic and environmental risk factors on an outcome of interest. When planning a study, alternate study designs are evaluated to minimize bias and achieve a large enough sample size from available resources. With the enormous volumes of high-quality biomedical data housed within its numerous biobanks, Norway is particularly well-suited to spearhead the investigation of a wide array of exposures and outcomes in a systematic manner. The rich array of longitudinal phenotypic data also permits an assessment of geneenvironment- timing interactions. Maximizing the research potential inherent in Norwegian biobanks is the overarching aim of Biobank Norway, an infrastructure project recently funded by the Norwegian Research Council. The development of advanced statistical tools for the analysis of high-throughput genomic data is critical to fulfill this aim and biostatistics platforms have been key elements of major biobank harmonization initiatives. However, many of these approaches have focused on traditional case-control designs. To exploit the particular advantages inherent in the Norwegian Mother and Child Cohort Study (MoBa), we describe here models to analyze the special data configurations available with offspring-parent designs. These models and the statistical tools outlined in this review were developed through the support of Biohealth Norway, a biobank platform funded by the Norwegian Functional Genomics Research Program (FUGE).

Comparison of global tests for functional gene sets in two-group designs and selection of potentially effect-causing genes

An important object in the analysis of high-throughput genomic data is to find an association between the expression profile of functional gene sets and the different levels of a group response. Instead of multiple testing procedures which focus on single genes, global tests are usually used to detect a group effect in an entire gene set. In a simulation study, we compare the power and computation times of four different approaches for global testing. The applicability of one of these methods to gene expression data is demonstrated for the first time. In addition, we propose an algorithm for the detection of those genes which might be responsible for a group effect. We could detect that the power of three of the approaches is comparable in many settings but considerable differences were detected in the computation times. Our proposed gene selection algorithm was able to detect potentially effect-causing genes in artificial sets with high power when many genes were altered with a small effect, while classical multiple testing was more powerful when few genes were altered with a large effect. An R-package called 'RepeatedHighDim' which implements our new global test procedures is made available from http://cran.r-project.org/.

Predicting genetic regulatory response using classification.

Studying gene regulatory mechanisms in simple model organisms through analysis of high-throughput genomic data has emerged as a central problem in computational biology. Most approaches in the literature have focused either on finding a few strong regulatory patterns or on learning descriptive models from training data. However, these approaches are not yet adequate for making accurate predictions about which genes will be up- or down-regulated in new or held-out experiments. By introducing a predictive methodology for this problem, we can use powerful tools from machine learning and assess the statistical significance of our predictions. We present a novel classification-based method for learning to predict gene regulatory response. Our approach is motivated by the hypothesis that in simple organisms such as Saccharomyces cerevisiae, we can learn a decision rule for predicting whether a gene is up- or down-regulated in a particular experiment based on (1) the presence of binding site subsequences ('motifs') in the gene's regulatory region and (2) the expression levels of regulators such as transcription factors in the experiment ('parents'). Thus, our learning task integrates two qualitatively different data sources: genome-wide cDNA microarray data across multiple perturbation and mutant experiments along with motif profile data from regulatory sequences. We convert the regression task of predicting real-valued gene expression measurements to a classification task of predicting +1 and -1 labels, corresponding to up- and down-regulation beyond the levels of biological and measurement noise in microarray measurements. The learning algorithm employed is boosting with a margin-based generalization of decision trees, alternating decision trees. This large-margin classifier is sufficiently flexible to allow complex logical functions, yet sufficiently simple to give insight into the combinatorial mechanisms of gene regulation. We observe encouraging prediction accuracy on experiments based on the Gasch S.cerevisiae dataset, and we show that we can accurately predict up- and down-regulation on held-out experiments. We also show how to extract significant regulators, motifs and motif-regulator pairs from the learned models for various stress responses. Our method thus provides predictive hypotheses, suggests biological experiments, and provides interpretable insight into the structure of genetic regulatory networks. The MLJava package is available upon request to the authors. Supplementary: Additional results are available from http://www.cs.columbia.edu/compbio/geneclass