Systematic Integration of GWAS with Co-Expression Networks to Detect Causal Genes for Elemental Accumulation in Zea Mays and Arabidopsis thaliana Using Camoco

Genome wide association studies (GWAS) have become the de facto approach to identify agriculturally important genes. However, complex traits such as elemental accumulation are often associated with potentially hundreds of statistically significant SNPs, many of which are located outside known gene models. Identifying causal genes for complex traits quickly become unwieldy even using straightforward marker to gene mapping such as the closest flanking gene.

Here, we present a computational framework called Camoco (Co-analysis of molecular components)  that integrates loci identified by GWAS with co-expression networks to identify a focused set of candidate loci with functional coherence. This framework analyses the overlap between candidate loci generated from GWAS and the co-expression interactions that occur between them. Camoco implements a suite of functions that perform SNP-to-gene mapping, builds and functionally validates co-expression networks, and provides a robust bootstrapping model to evaluate the statistical significance of candidate loci.

Using Camoco, candidate genes were identified from GWAS measuring elemental accumulation in maize seeds as well as arabidopsis leaves, seeds and roots. Three maize and four arabidopsis co-expression networks were generated using diverse genotypic as well as tissue based gene expression data. On average, candidate gene lists identified by GWAS were reduced by two orders of magnitude by integrating co-expression network information, which produced a focused set of candidates with both strong associations with the phenotype as well as evidence for functional coherence in the co-expression network. Identifying candidates is an important step to resolve genes responsible for elemental accumulation and other agriculturally important traits.