Genome-wide trait analysis usually aim to analyze complex trait by association mapping, which essentially deploy a statistical method for the variance component estimation and quantifying the narrow sense contribution(additive effect only) of a trait' heritability of a particular subset of genetic variant(typically refer to SNPs). In short, Genome wide trait analysis fundamentally is an association mapping between the genotypic sequence variation and phenotypic variation.

This association method however can be extended to other types of associative omics studies (not only the genome-wide association studies or 'GWAS'), and generate a broad sense 'Ome'-wide association mappings, such as transcriptome wide gene expression variation to phenotypic trait variation mapping (TWAS or Associaitive Transcriptomics), or metabolome wide metabolite abundance variation to phenotypic trait variation mapping (MWAS or Associative Metabolomics)

We developed a new linear mixed mode, which is capable of dissecting the whole estimated variance into mulitple components: e.g. additive effect and the marker-marker interaction as additive-additive. This model can essentially map QTL by controlling the polygenic background effects.

Based on this linear mixed model, we developed PATOWAS for analyzing trait through 'Ome'-wide association studies. The PATOWAS is composed of two primary sub-pipelines. Sub-pipeline 1 consist of one module only, namely 'km_cal', which is designed for kinship matrix calculation, and Sub-pipeline 2 is designed for association mapping and integration of three related analysis modules: (a) module one, namely, 'vc_anal' for the three variance component analysis, (b) moudle two, namely 'ps_main' for 1D P-value scanning for mian addtive effect, and (c) module three, namely, 'ps_inter' for 2D P-value scanning for interaction effect.

PATOWAS allows users to run only a portion of the pipeline according to the input data and user-configured parameter(e.g., users can perform only kinship matrix calculations and the three variance component analysis, or user can perform only kinship matrix calculations.) Such configuration flexibility enables users to tuilize PATOWAS to generate specific data, such as kinship matrix, and use it in their own statistical analysis.

Once data analyzed, PATOWAS returns the corresponding kinship matrix Ka, Kaa, the estimated mutiple variance component , main effect as one dimensional (1D) P-value vector , and interaction effect as two dimensional (2D) P-value matrix. Based on these mapping results, the associative omics mapping, or broad sense 'Ome' wide association studies can be further conducted.

Reference:

1. Yang, Jian, et al. "GCTA: a tool for genome-wide complex trait analysis." The American Journal of Human Genetics 88.1 (2011): 76-82.

2. Bradbury, Peter J., et al. "TASSEL: software for association mapping of complex traits in diverse samples." Bioinformatics 23.19 (2007): 2633-2635.

3. Xu, S., Mapping Quantitative Trait Loci by Controlling Polygenic Background Effects. Genetics, 2013. 195(4):p.1709-23.

4. Zhang W, Dai X, Wang Q, Xu S, Zhao PX, PEPIS: A Pipeline for Estimating Epistatic Effects in Quantitative Trait Locus Mapping and Genome-Wide Association Studies, 2016. PLoS Comput Biol, 12(5).

5. Zhang, W., Dai, X., Xu, S., & Zhao, P. X. (2018). 2D association and integrative omics analysis in rice provides systems biology view in trait analysis. Communications Biology, 1, 153. http://doi.org/10.1038/s42003-018-0159-7.