|Other Abstract||Selective breeding for important economic traits of Pacific white shrimp Litopenaeus vannamei (L. vannamei) is the foundation of shrimp aquaculture industry. During the past decade, the genetic improvement of important economic traits has made remarkable achievements by using traditional selective breeding approaches. However, the longtime breeding interval, high costs and low selection accuracy restricted the further development of selective breeding in L. vannamei. Therefore, it is extremely valuable to introduce new approaches to accelerate the process of genetic improvement of important economic traits in L. vannamei. Recently, the methods based on molecular markers, including genome-wide association study (GWAS) and genomic selection (GS), have been turned out to be promising candidate for the future application of selective breeding. In the present study, GWAS and GS has been evaluated in the breeding of L. vannamei. The main progresses achieved in this thesis are as follows:|
The potential factors that may influence the accuracy of GS in L. vannamei were investigated by different design of cross-validation in a multi-family population. At first, the heritability for L. vannamei based on the full set of markers (23K) was estimated to be 0.321 for body weight, and 0.452 for body length. The estimated heritability increased rapidly with the increase of the marker density from 0.05K to 3.2K, and then it tended to be stable for both traits. For genomic prediction on the growth traits of L. vannamei, three statistic models (RR-BLUP, BayesA and Bayesian LASSO) showed similar performance for the prediction accuracy of genomic estimated breeding value (GEBV). The prediction accuracy will be increased with the increasing of marker density. However, the marker density would bring a weak effect on the prediction accuracy when the marker number increased to 3.2K. In addition, genetic distance can influence the GS accuracy significantly. A distant genetic relationship between reference and validation population would result in a poor performance of genomic prediction for growth traits in L. vannamei.
GWAS has been used to detect the markers associated with body length and body weight based on the analyses of 3 960 SNPs in a full-sib family of L. vannamei The results showed that 52 SNPs were significantly associated with body length and 47 SNPs associated with body weight. Through further gene annotation and validation, we found that the SNP marker (M1286-15-3) located on protein kinase C (PKC) was significantly associated with body weight. The average body weight of individuals with CC genotype was significantly higher than that of the individuals with CT and TT genotype. The highest expression of PKC gene in the muscle of L. vannamei suggested that PKC might be an important gene in regulating the growth of shrimp.
Through GWAS analyses, SNP markers associated with 10 growth-related traits were identified using a 2b-RAD genotyping platform based on a multi-family population These markers located on many linkage groups which suggested that the growth traits of L. vannamei were controlled by minor-polygenes. Besides, linkage disequilibrium (LD) analysis based on the multi-family population showed that the decay of LD was very fast. When the distance between markers was 18 kb, the value of r2 was 0.2. Considersing the average distance between adjacent markers was 226.12 kb in the current study, the marker density must be improved to increase the reliability of GWAS for the important economic traits of L. vannamei.
The disease-resistance traits of L. vannamei were analyzed by GWAS to identify the related markers based on the resistant group and sensitve groups to Vibrio parahaemolyticus (V. parahaemolyticus) obtained by challenge test. The results showed that 8 SNPs were significantly associated with the resistance of L. vannamei to V. parahaemolyticus at genome-wide significance level with P<0.001 , and 187 SNPs were associated with the resistance at P<0.01. Through the positioning analysis of SNPs on the linkage groups, 108 SNPs associated with the resistance to V. parahaemolyticus at P<0.01 were mapped to 30 linkage groups, and the number of significantly associated SNPs in these linkage groups ranged from 1 to 9. These data gave us a hint that the resistance trait of L. vannamei to V. parahaemolyticus is a quantitative trait which was controled by a lot of QTLs with small effects.
The performance of GS models was evaluated for the body length and body weight of L. vannamei using 3 960 SNPs genotyped in 205 individuals from a full-sib family. Three GS models (RR-BLUP, BayesA, and Bayesian LASSO) were used to calculate the GEBV, and their predictive ability was assessed by the reliability of the GEBV and the bias of the predicted phenotypes. The mean reliability of the GEBVs for body length and body weight predicted by different models was 0.296 and 0.411, respectively. For each trait, three models showed very similar performance with respect to the predictability. The regression coefficients estimated by three models were very close, Therefore, we suggested that three models appeared practicable when GS was applied in a L. vannamei population.