systems genetics approaches
v1.32
Short introduction & scheme of tools
mediation reverse-mediation PheWAS PheWAS ePheWAS
Identifying genetic and environmental factors that impact complex traits and common diseases is a high biomedical priority. These webtools make use of the multilayered datasets from the BXD mouse population to expedite in silico gene function prediction through a series of integrative and complimentary systems analytical approaches, including (expression-based) phenome-wide association, transcriptome-/proteome-wide association, and (reverse-) mediation analysis.
See more
News and updates
Nov 29, 2017
Our paper titled An integrated systems genetics and omics toolkit to probe gene function is published online in Cell Systems.
May 15, 2017
Mediation and reverse-mediation analysis are implemented.
May 1, 2017
ePheWAS, a way to identify associations between gene transcript or protein levels and phenome was updated.
See more
Statistical summary of data
Over the past decades, hundreds of studies on the BXD population have created a wealth of multi-layered omics data, ranging from genomic, transcriptomic, proteomic, metabolomic, to phenomic data.
See more

Identifying genetic and environmental factors that impact complex traits and common diseases is a high biomedical priority. The webtools make use of the multilayered datasets from the BXD mouse population to expedite in silico gene function prediction through a series of integrative and complimentary systems analytical approaches, including (expression-based) phenome-wide association, transcriptome-/proteome-wide association, and (reverse-) mediation analysis (Figure 1-2). Applied to large multi-omics datasets from the BXD mouse genetic reference population, novel associations between genes and clinical and molecular phenotypes, and regulatory relations between genes could be identified. Importantly, unlike many connections seen in classic loss-of-function studies, the gene-gene and gene-phenotype links identified in population studies were robust and translated well across populations and species.

Figure 1. Systems approaches that can be applied using the multi-omic data in BXDs. Particular approaches developed in this study are highlighted with red arrows.

Figure 2. Flowchart for the systems approaches using the multi-omic BXD data. The gene-of-interest is first inspected on three aspects in the BXD GRP, i.e. the existence of genetic variations, e(p)QTLs, and its expression pattern across strains. PheWAS can be applied on genes that possess high-impact variants or significant cis-QTLs to identify the associated traits. Genes that have cis- or trans-QTLs can be analyzed to reveal the regulatory mechanism of gene expressions through mediation analysis. Expression-based PheWAS (ePheWAS) investigates the association between gene expression and phenotypic traits.

Nov 29, 2017
Our paper titled An integrated systems genetics and omics toolkit to probe gene function is published online in Cell Systems.
May 15, 2017
Mediation and reverse-mediation analysis are implemented.
May 1, 2017
ePheWAS, a way to identify associations between gene transcript or protein levels and phenome was updated.
April 21, 2017
~5000 clinical phenotypes and transcriptome data from more than 30 tissues of the BXD population were added in our database.
April 21, 2017
PheWAS has been added in the webtools
April 1, 2017
The systems-genetics.org was built, and will be use to implement several systems genetics tools developed in our lab.

Over the past decades, hundreds of studies on the BXD population have created a wealth of multi-layered omics data, ranging from genomic, transcriptomic, proteomic, metabolomic, to phenomic data (Figure 1). All the data have been archived and are publicly available in GeneNetwork. We focus here on data from 93 BXD strains (including BXD1-BXD102), from the parental C57BL/6J and DBA/2J strains, and from reciprocal F1 hybrids (i.e. B6D2F1, D2B6F1), that collectively encompass the vast majority of all BXD data.

Figure 1. Overview of multi-omic data from the BXD population

Phenome —Since the first publication on the BXDs, well over 200 research groups have generated behavioral, neurological, pharmacological, immunological, and more recently, metabolic phenotypes for this family. Thanks to recent advances in high-throughput phenotyping, the size and variety of the BXD phenome has increased exponentially since 2010, to ~5,000 quantitative clinical phenotypes as of December 2016 (Figure 2, detailed list could be downloaded here ).

Figure 2. Number of phenotypes collected on the BXD population increases over years

Transcriptome, proteome, and metabolome — There are approximately 200 transcriptome datasets collected from 34 tissues of the BXD cohort (Figure 3, detailed list could be downloaded here ). Furthermore, ~2,600 liver proteins quantified by SWATH-MS, and ~980 metabolites measured in liver and muscle were also included in the analyses.

Figure 3. Graphical illustration of the tissues profiled at the transcriptomic levels from the BXDs.

In combination, we employed deep phenome data consisting of ~5,000 phenotypic traits, and more than 200 transcriptome, proteome, and metabolome datasets for the BXD population—by far the largest coherent multi-omics data assembled for any experimental cohort—as the foundation to identify the genetic architecture underlying complex traits and diseases.

Table 1. Summary of BXD multi-omics datasets