Study Of The Algorithm For Genome-wide Association Analyses Based On Linear Mixed Model

Genome-wide association study(GWAS)has become an important approach for analyzing the genetic mechanism of complex traits in animals and plants.Currently,linear mixed model(LMM)is the major method for GWAS with excellent performance for adjusting population structure and individual relatedness.However,with the increasing of sample size and marker density,LMM methods are confronted with two major limitations.First,it is difficult to shrinkage the nonsense marker effect,which often increasing the false positive rate because the current LMM methods assume the marker effect as the fixed effect.Second,the current LMM methods estimate variance component and marker effect continually,which requires long time for calculation.To solve the limitations above,we applied composite interval mapping(CIM)strategy into LMM method(CIM polygene model).In addition,we used two flanking markers to eliminate interference from two sides of the target marker.We used CIM polygene model and other two common LMM methods to identify the candidate genes that associated with carcass weight and bone weight in Chinese Simmental cattle.Then,the performance of these three methods was assessed based on GWAS results.Moreover,to improve the operation efficiency of GWAS,we applied score test in LMM method(LMM-Score).We analyzed simulation data with the common LMM method and LMM-Score.Also,we applied LMM-Score to identify candidate genes for Pure meat weight(PMW),Fore Shank weight(FSW)and Silverside weight(SSW)in Chinese Simmental cattle.Our results showed that the CIM polygene model performed better than common LMM,this method identified more candidate genes associated with carcass weight and bone weight which including GCNT4,ALDH1A2,LCORL and WDFY3.Also,the LMM-Score substantially increased the computational speed and reliability when compared with common LMM.We also identified SOX17,RP1,LYN,RPS20,snoU54,U1,MOS,PLAG1,CHCHD7 and SDR16C5 as the novel candidate genes that associated with PMW,FSW and SSW.Our study offer several new insights for LMM GWAS methods and our finding further help to explore the genetic mechanism of carcass traits in Chinese Simmental cattle.