The Demographic History And Selection Mode During Domestication In Soybean

Selection is a major driving force during the adaptive process of species,and identification of loci subjected to selection is a main content of population genetics.In the past 20 years,efforts on selection detection were largely focused on a hard selective sweep model,while in recent years,soft sweep model has obtained increasing interest for their potential importance on the adaptations of wild species.So far,few if any studies on the role of soft sweeps in the soybean domestication and the adaptation process of wild soybean have been reported.First,we examined the genomic selection signatures of a soft sweep gene Dt1.Dt1 is the first domestication-related gene identified at the molecular level in soybean,and also provides us a gene case of soft sweep.Artificial selection on four nonsynonymous mutations in the coding region of Dt1 converted indeterminant growth habit in wild soybean to determinant growth habit in cultivated populations.To better understand the signatures of the soft sweeps across the Dt1 locus and track the origins of the determinate alleles,we examined patterns of nucleotide variation in Dt1 and the surrounding genomic region of approximately 800 kb with 216 cultivated and 147 wild soybeans.Four local,asymmetrical hard sweeps on four determinate alleles,sized approximately 660,120,220and 150 kb,were identified,which constitute the soft sweeps for the adaptation.Haplotype network and phylogenetic analysis showed that the four determinant haplotypes had three independent origins,one of which was directly from wild soybean,and the other three were from indeterminant genotypes of cultivated soybean.Then,based on previously-published resequencing data of 302 soybean accessions(62 wild soybean,130 landraces,and 110 improved varieties),we analyzed the changes of effective population size of cultivated and wild soybean in the past one million years,constructed their neutral demographic models,and demonstrated the relative contribution of hard and soft selective sweeps in domestication of cultivated soybean and adaptations of wild soybean.The history of effective population size(Ne)in the past one million years was analyzed for wild and cultivated soybean by PSMC and SMC++methods.The Ne of cultivated soybean began to decline about 24000 years ago,and it was completely separated from wild soybean about 4000 years ago,and the Ne continued to decline to a minimum size of?4000 until 1500 years ago.The statistics estimated from coalescent simulation based on the inferred pattern of Ne above were close to that from real resequencing data,indicating a robust and convincing inference for the population history during the soybean domestication.Subsequently,we employed H12 statistic to scan for the selective candidates in wild and cultivated soybean populations with 7 increasing window size,and then calculated the relative proportions of hard and soft sweep genes by H2/H1.The results showed soft sweeps are much more prevalent in the wild genome,while hard selective sweeps were the dominant mode during the domestication of cultivated soybean.Further analysis for the minor part of soft-sweep genes showed that most of those soft signatures were actually caused by the stepwise hard sweeps,which further confirmed the dominant role of hard sweeps during soybean domestication.At last,taking six domestication-related genes including Dt1 as study object,we evaluated the performance of five statistics based on different selection characteristics and performed genome-wide scan for the selection targets based on the common characteristics of the domestication-related genes.Performance of various test statistics suggests that differentiation-based methods are robust for detecting complete hard sweeps,and that LD-based strategies perform well for identifying recent/ongoing sweeps;however,none of the test statistics detected a known soft sweep we previously documented at the domestication gene Dt1.Genome scans yielded a set of 66 candidate loci that were identified by both differentiation-based(F_ST)and LD-based(i HH)methods;notably,this shared set overlaps with many previously identified QTLs for soybean domestication/improvement traits.This provides important clues for the follow-up study of their biological functions.This study is expected to provide important insights on the selection modes and molecular mechanism for soybean and even crop domestication from a theoretical perspective;and from the perspective of practice,it will provide a solid theoretical basis for further excavation of key genes controlling important traits of soybean and effective cultivation of soybean excellent soybean varieties.