Joint analysis of multiple phenotypes: summary of results and discussions from the Genetic Analysis Workshop 19

Abstract

For Genetic Analysis Workshop 19, 2 extensive data sets were provided, including whole genome and whole exome sequence data, gene expression data, and longitudinal blood pressure outcomes, together with nongenetic covariates. These data sets gave researchers the chance to investigate different aspects of more complex relationships within the data, and the contributions in our working group focused on statistical methods for the joint analysis of multiple phenotypes, which is part of the research field of data integration. The analysis of data from different sources poses challenges to researchers but provides the opportunity to model the real-life situation more realistically.Our 4 contributions all used the provided real data to identify genetic predictors for blood pressure. In the contributions, novel multivariate rare variant tests, copula models, structural equation models and a sparse matrix representation variable selection approach were applied. Each of these statistical models can be used to investigate specific hypothesized relationships, which are described together with their biological assumptions.The results showed that all methods are ready for application on a genome-wide scale and can be used or extended to include multiple omics data sets. The results provide potentially interesting genetic targets for future investigation and replication. Furthermore, all contributions demonstrated that the analysis of complex data sets could benefit from modeling correlated phenotypes jointly as well as by adding further bioinformatics information.

Fig. 1

Results of PubMed literature search. Results of a literature search on PubMed on June 25, 2015, for articles published between January 1, 1990 and June 1, 2015, containing any of “(data integration OR joint model OR joint analysis OR multiple phenotypes OR multivariate model OR multivariate statistics)” as well as “(omics OR NGS OR high-throughput OR xx)”, where xx was any permutation of 2 omics measures from genomics, transcriptomics, epigenomics, proteomics, and metabolomics, each parameterized by different possible keywords. The number of retrieved articles in 2015 (left panel) is multiplied by 2.4 to predict published articles in 2015. The top 20 journals with the most publications are shown in the right panel

Fig. 2

Simplified sketches of the underlying biological models and assumptions of the working group papers. BP _i, BP measure at the i ^th visit; DBP, diastolic blood pressure; GE, gene expression; LBP, latent variable affecting both systolic and diastolic blood pressure; MURAT, multivariate rare-variant association test; SBP, systolic blood pressure; SEM, structural equation modeling; SNV, single nucleotide variant; SRVS, sparse representation-based variable selection. For a more detailed presentation of the models, please refer to the original articles