Integration of Multi-Omics Datasets

Omics approaches, including genomics, transcriptomics, epigenomics, proteomics and metabolomics, offer an efficient and unbiased way to more comprehensively characterize biological pathways over mechanistic studies focused on individual genes and proteins. These high-throughput technologies provide researchers with a snapshot of biological systems, aid in drug development and increase our understanding of disease mechanisms. Early successes in the use of omics technologies, and the growing number of available omics datasets, have led to optimism about the future of precision medicine, in which diagnostic tests and treatments are increasingly tailored to individuals’ genetic, environmental and lifestyle factors. We have successfully applied omics approaches to study traits related to chronic pulmonary diseases, including the effects of glucocorticoids and bronchodilators on asthma tissues. Two reviews of omics approaches are:

As an example of a single omics modality study, we performed the largest airway smooth muscle (ASM) glucocorticoid response transcriptomic study to date and found that at the level of transcription, fatal asthma and non-asthma donor-derived ASM responded strongly and similarly to exposure to the glucocorticoid budesonide. This finding was surprising because ASM cells derived from fatal asthma donors retain differences in proliferative and contractile outcomes that suggest their response to glucocorticoids differs from that of donors without asthma. By comparing differential expression results of 14 publicly available glucocorticoid response transcriptomic datasets for 7 cell types, we also found a distinct ASM glucocorticoid response signature, underscoring the importance of obtaining omics datasets for disease-related tissues. More details are in this paper:

We have developed bioinformatics pipelines for analysis of transcriptomic (i.e., microarray and RNA-Seq gene expression studies) and epigenomic (i.e., ChIP-Seq) data to increase reproducibility of results and facilitate use of these approaches by investigators who are less familiar with them. Because of our long-standing focus on pulmonary traits, we are sought by collaborators for whom our familiarity with disease- and tissue-specific issues is an asset in the design studies, and analysis and interpretation of omics results.

Beyond analyzing single-modality omics data, we have developed and applied strategies to perform integrative analyses of omics data. This is a challenging area that requires creative yet rigorous analytical approaches, based on a thorough understanding of heterogeneous data types, use of computational resources and expertise to store and analyze large datasets. Examples include combining mouse and human association data, gene expression, and expression quantitative trait loci (eQTL) data in novel ways to aid in identifying disease-associated genes, and integration of transcriptomic and ChIP-Seq data to better understand asthma genetic associations.

Mining publicly available omics datasets that contain valuable and unused information has the potential to improve our understanding of complex diseases through the careful integration of datasets to obtain more comprehensive views of genomic relationships and help to prioritize genes for functional validation studies. For pulmonary disease research, it is very helpful to have cell type-specific information, rather than that based on whole lung, which is often used for convenience. Using asthma as a disease model, we have integrated results from individual expression experiments in which specific tissues, with intact genomes and in the context of knocked-down or overexpressed genes, were stimulated with pro-inflammatory cytokines, environmental pollutants, allergens, and/or drugs used for the treatment of asthma.