| Summary: | 博士 === 國立中興大學 === 動物科學系所 === 106 === Quantitative trait locus (QTL) detection is a classical approach to better understand the genetic architecture of complex traits and unravel genomic regions which are controlling quantitative variation of traits. The publication of the chicken genome sequence has made available, thousands of SNP that can be genotyped in an appropriate experimental design to detect with high accuracy genes located in chromosomal regions that control quantitative variations.
This dissertation is the result of a co-supervision program between France (AgroParisTech, APT, and Institut National de la Recherche Agronomique, INRA) and Taiwan (National Chung Hsing University, NCHU). A QTL detection project was set up in the real conditions of the humid sub-tropical climate in Taiwan, using an F2 population of 743 individuals produced by crossing the Taiwan Country chicken L2 line with an experimental line of Rhode Island Red layer R- provided by INRA in 2003. To assess adaptation to sub-tropical climate, a set of relevant traits was recorded including growth, immune response, egg production, egg quality, residual feed consumption, body composition, and meat quality. The analysis aimed at identifying QTL that control performance under humid sub-tropical conditions in chickens. Indeed, QTL regions affecting growth-related traits can be used as indicators of adaptation to sub-optimal conditions. Two methods were applied: genome-wide association study (GWAS with the GEMMA software) and interval mapping combining linkage disequilibrium (LD) and linkage analysis (LA) with QTLMap software (LDLA option). The position of the significant SNPs that flanked a QTL region was used to establish the correspondence between the results of QTL mapping and those of GWAS. Whole genome sequence data previously obtained for a few individuals of the L2 and R- lines were aligned on the GalGal4 reference and used to document the polymorphisms of the candidate genes that were chosen according to their positions and known functions.
There were few QTL regions detected by both methods, and the LDLA mapping analysis revealed a higher number of QTL regions as compared to GWAS. This was interpreted by a major contribution of the family information to QTL detection in an F2 design as compared to the LD information. It was also shown that QTL detection with GWAS without including the genomic relationship matrix led to many false positives. The whole-genome QTL analysis led to the identification of 112 QTL that corresponded to 100 non-overlapping regions which may influence adaptation of chickens to varying environmental conditions. Among the 112, 38 exhibited genome-wide significance. A particularly strong QTL was identified on the Z chromosome and affected all growth traits (body weight at different ages and shank length at 8 weeks of age). It could be exploited further by a crossbreeding scheme between R- males and L2 females. Most of the QTL regions detected for growth traits overlapped with previously published QTL, which suggests that these QTL have an effect across a range of environmental conditions and could be particularly useful for selection. Candidate genes could be proposed, such as the PTGER4, OSMR, or FGF10 genes which exhibited several SNPs between the L2 and R- sequenced individuals, including missense mutations. These mutations should be genotyped in the F2 as well as in the current lines. Genome-wide QTLs were identified for egg shape and egg quality traits but very few on egg number, with only one genome-wide QTL on pause length. Candidate genes such as the PRL and GDF9 genes were found for two yolk weight QTLs, the GDF9 gene may be of particular interest for further study.
|
|---|
