Page 175 - 《水产学报》2023年第1期

刘杨，等 水产学报, 2023, 47(1): 019612

Genomic prediction accuracy analysis of AHPND resistance genome prediction in
Litopenaeus vannamei using SNP panels with different densities

2
2
2
2
1,2
2
LIU Yang , LUAN Sheng , LIU Mianyu , LI Xupeng , MENG Xianhong , LUO Kun ,
2
2
2
2
2
SUI Juan , TAN Jian , DAI Ping , CAO Jiawang , CHEN Baolong , KONG Jie 2*
(1. College of Marine Science and Fisheries, Huzhou University, Huzhou 313000, China;
2. Key Laboratory for Sustainable Development of Marine Fisheries, Ministry of Agriculture and Rural Affairs, Laboratory for Marine
Fisheries Science and Food Production Processes, Pilot National Laboratory for Marine Science and Technology (Qingdao),
Yellow Sea Fisheries Research Institute, Chinese Academy of Fishery Sciences, Qingdao 266071, China)
Abstract: To evaluate the influence of SNP panels with different densities on the accuracy of genomic prediction
of AHPND resistance of Litopenaeus vannamei, 26 full-sib families were infected with Vp , and survival time
AHPND
of 686 individuals were recorded. A total of 242 individuals were genotyped by the liquid chip "Yellow Sea Chip
No.1" (55.0 K SNP), and the genetic parameters of survival time after Vp infection were estimated using the
AHPND
A, G and H relationship matrices. Eight low-density SNP panels (40.0 K, 30.0 K, 20.0 K, 10.0 K, 5.0 K, 1.0 K, 0.5
K, 0.1 K) were generated based on the 55.0 K SNP panel. GBLUP and ssGBLUP methods were used to predict the
GEBV of survival time, and prediction accuracy was calculated by cross-validation method and compared with
BLUP method. The estimated heritabilities of survival time after Vp infection were high (0.68-0.79). At 55.0
AHPND
K SNP density, the prediction accuracies obtained by BLUP, GBLUP and ssGBLUP were 0.424, 0.450 and 0.452,
respectively, in the data set with 242 genotyped individuals (G242). The accuracy of GBLUP and ssGBLUP was
6.13% and 6.60% higher than BLUP, respectively. For the data set with 686 phenotyped individuals (P686), the
prediction accuracies obtained by BLUP and ssGBLUP methods were 0.510 and 0.535, respectively. ssGBLUP
showed 4.90% greater accuracy than BLUP. For 8 low-density SNP panels, genomic prediction accuracies were all
low (1.1%-1.8%) in G242 and P686 data sets when the SNP density was greater than or equal to 10.0 K. With the
decrease of SNP density since 10.0 K, the accuracy of genomic prediction also decreased in the two data sets, The
density of 5.0 K decreased by 0.6%-2.6%, 1.0 K decreased by 5.8% -11.0%, 0.5 K decreased by 11.4%-17.2%, and
0.1 K decreased by 38.8%-41.6%. The correlation coefficients of genomic relationship coefficient and GEBV
obtained using 10.0 K and 55.0 K SNP panels were all higher than 0.99, which indicates that the relationship
between sibling individuals and their GEBVs can be accurately predicted by using the 10.0 K SNP panel. The res-
ults showed that the 10.0 K SNP panel could replace 55.0 K SNP panel to perform genetic evaluation for survival
time after Vp AHPND infection,which provided a reference for the design of low density SNP chips.
Key words: Litopenaeus vannamei; AHPND; low-density SNP panel; accuracy of genomic prediction
Corresponding author: KONG Jie. E-mail: Kongjie@ysfri.ac.cn
Funding projects: National Key Research and Development Program of China (2022YFD2400202); Science and
Technology Innovation Team Project of Chinese Academy of Fishery Sciences (2020TD26); Scientific Research
Project of Academician Innovation platform in Hainan Province (YSPTZX202104); Litopenaeus vannamei Breed-
ing Center of the Hengxing (2021E05032); China Agriculture Research System (CARS-48)

https://www.china-fishery.cn 中国水产学会主办 sponsored by China Society of Fisheries
10