Meta-analysis approaches to combine multiple gene set enrichment studies

From SNP to pathway-based GWAS meta-analysis: do current meta-analysis approaches resolve power and replication in genetic association studies?

Genome-wide association studies (GWAS) have benefited greatly from enhanced high-throughput technology in recent decades. GWAS meta-analysis has become increasingly popular to highlight the genetic architecture of complex traits, informing about the replicability and variability of effect estimations across human ancestries. A wealth of GWAS meta-analysis methodologies have been developed depending on the input data and the outcome information of interest. We present a survey of current approaches from SNP to pathway-based meta-analysis by acknowledging the range of resources and methodologies in the field, and we provide a comprehensive review of different categories of Genome-Wide Meta-analysis methods employed. These methods highlight different levels at which GWAS meta-analysis may be done, including Single Nucleotide Polymorphisms, Genes and Pathways, for which we describe their framework outline. We also discuss the strengths and pitfalls of each approach and make suggestions regarding each of them.

Dissecting Meta-Analysis in GWAS Era: Bayesian Framework for Gene/Subnetwork-Specific Meta-Analysis

Over the past decades, advanced high-throughput technologies have continuously contributed to genome-wide association studies (GWASs). GWAS meta-analysis has been increasingly adopted, has cross-ancestry replicability, and has power to illuminate the genetic architecture of complex traits, informing about the reliability of estimation effects and their variability across human ancestries. However, detecting genetic variants that have low disease risk still poses a challenge. Designing a meta-analysis approach that combines the effect of various SNPs within genes or genes within pathways from multiple independent population GWASs may be helpful in identifying associations with small effect sizes and increasing the association power. Here, we proposed ancMETA, a Bayesian graph-based framework, to perform the gene/pathway-specific meta-analysis by combining the effect size of multiple SNPs within genes, and genes within subnetwork/pathways across multiple independent population GWASs to deconvolute the interactions between genes underlying the pathogenesis of complex diseases across human populations. We assessed the proposed framework on simulated datasets, and the results show that the proposed model holds promise for increasing statistical power for meta-analysis of genetic variants underlying the pathogenesis of complex diseases. To illustrate the proposed meta-analysis framework, we leverage seven different European bipolar disorder (BD) cohorts, and we identify variants in the angiotensinogen (AGT) gene to be significantly associated with BD across all 7 studies. We detect a commonly significant BD-specific subnetwork with the ESR1 gene as the main hub of a subnetwork, associated with neurotrophin signaling (p = 4e⁻¹⁴) and myometrial relaxation and contraction (p = 3e⁻⁰⁸) pathways. ancMETA provides a new contribution to post-GWAS methodologies and holds promise for comprehensively examining interactions between genes underlying the pathogenesis of genetic diseases and also underlying ethnic differences.

The microbiome, genetics, and gastrointestinal neoplasms: the evolving field of molecular pathological epidemiology to analyze the tumor-immune-microbiome interaction

Metagenomic studies using next-generation sequencing technologies have revealed rich human intestinal microbiome, which likely influence host immunity and health conditions including cancer. Evidence indicates a biological link between altered microbiome and cancers in the digestive system. Escherichia coli and Bacteroides fragilis have been found to be enriched in colorectal mucosal tissues from patients with familial adenomatous polyposis that is caused by germline APC mutations. In addition, recent studies have found enrichment of certain oral bacteria, viruses, and fungi in tumor tissue and fecal specimens from patients with gastrointestinal cancer. An integrative approach is required to elucidate the role of microorganisms in the pathogenic process of gastrointestinal cancers, which develop through the accumulation of somatic genetic and epigenetic alterations in neoplastic cells, influenced by host genetic variations, immunity, microbiome, and environmental exposures. The transdisciplinary field of molecular pathological epidemiology (MPE) offers research frameworks to link germline genetics and environmental factors (including diet, lifestyle, and pharmacological factors) to pathologic phenotypes. The integration of microbiology into the MPE model (microbiology-MPE) can contribute to better understanding of the interactive role of environment, tumor cells, immune cells, and microbiome in various diseases. We review major clinical and experimental studies on the microbiome, and describe emerging evidence from the microbiology-MPE research in gastrointestinal cancers. Together with basic experimental research, this new research paradigm can help us to develop new prevention and treatment strategies for gastrointestinal cancers through targeting of the microbiome.

Pathway-based meta-analysis for partially paired transcriptomics analysis

Pathway-based differential expression analysis allows the incorporation of biological domain knowledge into transcriptomics analysis to enhance our understanding of disease mechanisms. To integrate information among multiple studies at the pathway level, pathway-based meta-analysis can be performed. Paired or partially paired samples are common in biomedical research. However, there are currently no existing pathway-based meta-analysis methods appropriate for paired or partially paired study designs. In this study, we developed a pathway-based meta-analysis approach for paired or partially paired samples. Meta-analysis on the transcriptomics profiles were conducted using p-value-based, rank-based, and effect size-based algorithms. The application of our approach was demonstrated using partially paired data from psoriasis transcriptomics studies. Upon combining six transcriptomics studies, genes related to the cell cycle and DNA replication pathways are found to be highly perturbed in psoriatic lesional skin samples. Results were validated externally with independent RNA-Seq data. Comparison with existing pathway meta-analysis methods revealed consistent results, with our method showing higher detection power. This study demonstrated the utility of our newly developed pathway-based meta-analysis that allows the incorporation of partially paired or paired samples. The proposed framework can be applied to omics data including but not limited to transcriptomics data.

Gene set meta-analysis with Quantitative Set Analysis for Gene Expression (QuSAGE)

Small sample sizes combined with high person-to-person variability can make it difficult to detect significant gene expression changes from transcriptional profiling studies. Subtle, but coordinated, gene expression changes may be detected using gene set analysis approaches. Meta-analysis is another approach to increase the power to detect biologically relevant changes by integrating information from multiple studies. Here, we present a framework that combines both approaches and allows for meta-analysis of gene sets. QuSAGE meta-analysis extends our previously published QuSAGE framework, which offers several advantages for gene set analysis, including fully accounting for gene-gene correlations and quantifying gene set activity as a full probability density function. Application of QuSAGE meta-analysis to influenza vaccination response shows it can detect significant activity that is not apparent in individual studies.