Simplified Method Of Data Simulating For Genome Selection And Development Of Network Breeding Analysis And Evaluation System Of Shellfish

This research includes two parts: One is a simplified way of simulating data for a genome-wide selection analysis; and the other is the developing of Analysis and Evaluation System for Molluscan Genetical Breeding.Currently the data applied for a Genome-Wide Selection analysis are usually simulated for all the information of whole genome. This approach is very slow when simulating a large and complex genome and the combination of information of multi-chromosome is very time-consuming. Although we can use higher-performance computers to run the simulation programs to speed up the simulation process, the higher cost is not so worthy. In order to improve this problem, this study discusses a simplified simulation method for simulating whole genome information—using the information of a single pair of chromosomes to infer the whole one. In theory, the information of one pair of chromosomes can be used to infer a complete genome: the simulation of information of a pair of chromosomes with the heritability of 0.03 is equivalent to a simulation that simulates information for 10 pairs of chromosomes with the heritability of 0.30 each, and somehow combine them together. This test simulates the information of a pair of chromosomes in one case. By repeating simulation and splicing the information of chromosomes, the result is used as a simulation case of a multi-pair of chromosomes. Data are simulated with different combination of multiple parameters: number of pairs of chromosomes, marker density, QTL density and heritability, and then the simulated data are applied for a whole genome selection analysis using GWBLUP and Bayes-C methods. The differences of simulating information of only one pair of chromosomes and of the whole genome-wide information are to be discussed. The results showed that the estimated breeding values became more accurate if there were more markers, higher heritability and less QTLs. Simulating the information of only one pair of chromosomes to apply for a genome-wide selection analysis does save a lot of time and memory storage of processing, and the loss of accuracy is not significant. So simulating the information of a single pair of chromosomes instead of the whole genome for testing new algorithmic methods is technically feasible.Scallop mariculture is one of the leading industries of sea-farming in China, and the scale of our scallop mariculture is the largest in the world. But with the rapid development of the mariculture, the lack of fine varieties, diseases and the scattered small private business of scallop aquaculture have become serious problems. Therefore, the cultivation of fine varieties has become one of the key factors for sustainable development of aquaculture. With the development of shellfish breeding and the accumulation of molecular marker information and phenotype information, a planned family training and the identification of individuals provide sufficient conditions for the development of BLUP assisted scallop breeding. This thesis describes the designee of a web-based system for data recording with the functions of (co-)variance component estimation and using animal model BLUP to estimate breeding values for selection, with the cooperation of Shellfish Laboratory of Molecular Genetics and Breeding Technology, Ocean University of China.