Development Of EigenGWAS-based Method For Detecting Genomic Selection Signatures And Its Cloud Computing Platform

Survival of the fittest is the core idea of Darwinian evolution.Since the last century,in combination with the development of population genetics and genomics,the research related to the Darwin's theory of evolution has ushered a brand-new stage-scientists began to investigate the genomic loci that under natural selection in order to trace and dissect the history of the evolution and the complex genetic mechanisms of environmental adaptation.Traditional methods of detecting genomic signatures under natural selection,especially represented by the genetic differentiation index F_st,require a clear subgroup classification of the population of interest.However,a large discrepancy in cost between sequencing and phenotyping leads to a large amount of genomic data without corresponding phenotypes,traditional methods are,therefore,not able to directly search for the loci been selected without subpopulation labels.In 2016,a novel method,EigenGWAS,for finding loci under natural selection,was proposed.Combining eigen-analysis and genome-wide association study,EigenGWAS provides as an"unsupervised"method to search the selected loci for those population without subpopulation information.However,EigenGWAS is currently only applicable to outcrossing populations such as maize,cotton,human and animals,and cannot analyze such inbred populations like various types of self-pollination plants,which are common in crop breeding in particular.In addition,most of the computer software developed based on the EigenGWAS method require complex configuration on running environment and operation in command line,and the computational efficiency of the software also cannot adapt to the rapidly growing computational demands with the current surge in data volume.In this study,we develop the EigenGWAS method for inbred populations and integrated it with the original framework,making the new EigenGWAS method is able to find genomic loci under selection in general populations.Through simulations and real data analysis in Arabidopsis population,great tit population,dog population and Chinese Han populations,we validated the effectiveness of the EigenGWAS method in detecting genomic signatures under selection,and the detected loci under selection provides valuable references for related studies in various research fields such as ecology,human genetics,and breeding.Furthermore,a new EigenGWAS software was developed using a hybrid programming technology of C and R languages.We further deployed the software on the cloud computing platform(www.eigengwas.com).The newly developed software has an user-friendly interface and abundant visualization functions on results,users are able to detecting the genomic loci under selection online by accessing the cloud platform through browser,without any software installation.The proposed method and the cloud computing platform developed in this study will greatly facilitate the work of related researchers and provide an easy-to-use,stable and efficient analytical tool in the related research fields.