Abstract 【Objective】In this study, the molecular breeding via genomic selection was carried out in the joint genomic evaluation on Yorkshire population in Beijing, predicting the breeding value of the new born boars and making selection, so as to improve the selection accuracy of breeding. 【Method】 An admixed population consisting of 4020 individuals from three Yorkshire breeding farms with different genetic background in Beijing was established as the reference group, and the reference animals were selected according to the performance testing records between 2007-2017 in those three pig farms. Three economic traits age at 100 kg (AGE), backfat thickness at 100 kg (BF) and total number born (TNB) were taken into account. The reference and candidate animals were genotyped with Illumina Porcine80K SNP chip. GEBV was estimated by single-step GBLUP (SSGBLUP) method which could make use of both pedigree information and genomic information. GEBVs of candidate boars on the growth traits and reproductive traits were predicted before castration and after performance testing, respectively. Afterwards, the elite candidates were selected according to their GEBVs. Meanwhile, the genetic connectedness among three pig farms was measured by connectedness rating.【Result】Our results showed that the genetic connectedness based on pedigree information among three Yorkshire breeding farms was too low to carry out traditional joint genetic evaluation. However,the genomic relationship coefficients of individuals between farms in G-matrix indicated that genetic links existed among different farms. The genomic selection could realize the joint genomic evaluation through establishing the genetic connectedness via genome-wide markers. A total of 1789 boars were genomic predicted.The accuracy of genomic prediction was largely improved, compared to traditional breeding methods. At the first time of implementing genomic selection or early selection (before the castration of boars), the accuracies of Pedigree Index (PI) for three traits, age at 100 kg (AGE), backfat thickness at 100 kg (BF) and total number born (TNB) were 0.55, 0.56 and 0.41, respectively. However, the accuracies of GEBV from genomic selection were increased to 0.65, 0.70 and 0.60 with improvement of 10, 14 and 19 percentage compared to PI selection, respectively. At the second time of implementing genomic selection (after performance testing), the accuracies of GEBV for AGE, BF and TNB were further increased to 0.78, 0.84 and 0.60, respectively, yielding 8, 12 and 19 percentage higher accuracy than EBV, respectively, in which the accuracies were 0.70, 0.72 and 0.41, respectively. The largest gain of genomic selection was on trait of TNB with low heritability. The early selection based on genomic selection had the same accuracy as traditional selection based on estimated breeding values calculated from performance testing, implying genomic selection could save breeding time and cost with keeping the same accuracy. The comparison of two implementations of genomic selection on 338 boars at different stage showed that the second genomic prediction after performance testing yielded higher accuracy, because the phenotypic records of these boars were also utilized. The accuracies of GEBV for AGE and BF were improved from 0.55, 0.62 to 0.72, 0.84 by increasing 17 and 22 percentage point, respectively. The unbiasedness coefficient was between 0.82 and 1.00, and the unbiasedness of GEBV on traits of AGE and BF were increased from 0.82 and 0.96 to 0.91 and 1.00, respectively. The lower unbiasedness of second genomic selection indicated that the reliability of selecting elite boars was higher.【Conclusion】 Genomic selection could establish genetic connectedness between different farms, enabling joint genetic evaluation which was not feasible in traditional breeding plausible and more breeding farms involved. Compared to traditional PI or EBV selection, genomic selection generated much higher accuracy, and the greatest improvement was obtained on the traits with low heritability. Genomic selection was useful to achieve early selection and to improve the breeding efficiency. Keywords:genomic selection;Yorkshire;admixed population;joint breeding;early selection
1.1.1表型数据 本研究数据来源于北京地区3家国家生猪核心育种场,北京六马养猪科技有限公司(BJLM,简称北京六马)、北京养猪育种中心(BBSCB,简称养猪中心)和北京顺鑫农业发展集团有限公司(BJXD,简称顺鑫农业),达100 kg体重日龄(age at 100 kg live weight , AGE)、100 kg活体背膘厚(backfat adjusted to 100 kg , BF)、总产仔数(total number born , TNB)等3个性状2007—2017年场内性能测定记录,其中繁殖记录为54888条,生长记录为78540条。
1.1.2基因型数据 参考群体:来自上述3个场的4020头大白猪组成基因组选择的参考群体,本研究样品利用天根血液基因组DNA提取试剂盒提取试验猪血样,个体基因型由Illumina公司Porcine SNP 80KBeadchip芯片SNP分型得到,共包含68528个SNP位点。对芯片数据进行如下质控处理:
Table 1 表1 表1基因组联合育种大白猪参考群体和候选群体规模统计 Table 1Population size of reference and candidate population from three Yorkshire breeding farms in genomic joint breeding
PENG ZZ . Discussion on the relevant issues in the interpretation and implementation of the "National Swine Improvement Program"(Continued) Swine Industry Science , 2011(10):106-108.(in Chinese) URL [本文引用: 1]
ZHANGQ, DING XD, CHEN YS . Development and application of swine genetic evaluation system in China Chinese Journal of Animal Science, 2015,51(8):61-65.(in Chinese) URL [本文引用: 3]
ZHANG JX, ZHANG SY, QIU XT, GAOH, WANG CC, WANGY, ZHANGQ, WANG ZG, YANG HJ, DING XD . The genetic connectedness of Duroc, Landrace and Yorkshire pigs in China ActaVeterinaria et ZootechnicaSinica, 2017,48(9):1591-1601.(in Chinese) URL [本文引用: 2]
ZHANGZ, ZHANGQ, DING XD . Advances in genomic selection in domestic animals Chinese Sci Bull, 2011,56(26):2212-2222.(in Chinese) [本文引用: 1]
SCHAEFFER LR . Strategy for applying genome-wide selection in dairy cattle , 2006,123(4):381-391. [本文引用: 2]
FORNIS, AGUILARI, MISZTALI. Genomic relationships and biases in the evaluation of sow litter size 2010: . 2-6. [本文引用: 3]
SONG HL, ZHANG JX, JIANGY, GAOH, TANG SQ, MI SR, YUF, MENG QL, XIAOW, ZHANGQ, DING XD . Genomic prediction for growth and reproduction traits in pig using an admixed reference population , 2017,95(8):3415-3424. [本文引用: 2]
SAATCHIM, MCCLURE MC, MCKAY SD, ROLF MM, KIMJ, DECKER JE, TAXIS TM, CGAPPLE RH, RAMEY HR, NORTHCUTT SL, BAUCKS, WOODWARDB, DEKKERS JC, FERNANDO RL, SCHNABEL RD, GARRICK DJ . Accuracies of genomic breeding values in American Angus beef cattle using K-means clustering for cross-validation , 2011,43(1):40-43. DOI:10.1186/1297-9686-43-40URL [本文引用: 1]
LONGN, GIANOLAD, ROSAGJ, WEIGEL KA, AVENDANOS . Machine learning classification procedure for selecting SNPs in genomic selection: Application to Early Mortality in Broilers , 2007,124(6):377-389. DOI:10.1111/j.1439-0388.2007.00694.xURL [本文引用: 1]
RIEDELSHEIMERC, CZEDIKEYSENBERGA, GRIEDERC, LISECJ, TECHNOWF, SULPICER, ALTMANNT, STITTM, WILLMITZERL, MELCHINGER AE . Genomic and metabolic prediction of complex heterotic traits in hybrid maize , 2012,44(2):217-220. DOI:10.1038/ng.1033URL [本文引用: 1]
GUO TT, LI HH, YAN JB, TANG JH, LI JS, ZHANG ZW, ZHANG LY, WANG JK . Performance prediction of F1 hybrids between recombinant inbred lines derived from two elite maize inbred lines , 2013,126(1):189-201. DOI:10.1007/s00122-012-1973-9URL [本文引用: 1]
GUO LH, LU LH, ZHANGT, SHI XF, YUANK, WANGZ, YANG SW . Research Progress and Prospects of Genome-wide Selection Breeding in Plants Journal of Shanxi Agricultural Sciences.2015,43(11):1558-1562.(in Chinese) URL [本文引用: 1]
GUTIERREZ AP . Genomic selection for improvement of Atlantic salmon aquaculture: [Doctor of Philosophy] , 2015. [本文引用: 1]
ZHOUL, YANG HW, ZHAO ZK, YANG HJ, LIU JF . The Application of Genomic Selection in Chinese Pig Breeding Industry Chinese Journal of Animal Science.2018,54(3):4-8.(in Chinese) URL [本文引用: 1]
CHANG CC, CHOW CC, TELLIER LC, VATTIKUTIS, PURCELL SM, LEE JJ . Second-Generation PLINK: Rising to the challenge of larger and richer datasets , 2015,4(1):4-7. [本文引用: 1]
BROWNING BL, BROWNING SR . A unified approach to genotype imputation and haplotype-phase inference for large data sets of trios and unrelated individuals , 2009,84(2):210-223. DOI:10.1016/j.ajhg.2009.01.005URL [本文引用: 1]
MATHUR PK, SSLLIVAN BP, CHESNAIS JP . Measuring connectedness: concept and application to a large industry breeding program , 2002. [本文引用: 2]
MAYHUR PK, SULLIVANB, CHESNAISJ . Estimation of the degree of connectedness between herds or management groups in the Canadian swine population , 1999. [本文引用: 3]
ZHANGY, YANG YZ, ZHANG JX, ZHANG SY, QIU XT, TANG SQ, DING XD . Comparison of Breeding Value Estimation by DMU,GBS and Herdsman Chinese Journal of Animal Science, 2017,53(7):29-35.(in Chinese) URL [本文引用: 1]
BAI JY, YANG YB, LI GL . Application of GBS Breeding Analysis System in the Teaching of Livestock Breeding Animal Husbandry and Feed Science, 2018,39(03):97-100.(in Chinese) URL [本文引用: 1]
CHRISTENSEN OF, LUND MS . Genomic prediction when some animals are not genotyped , 2010,42(1):2. DOI:10.1186/1297-9686-42-2 [本文引用: 2]
CHRISTENSEN OF, MADSENP, NIELSENB, OSTERSENT, SUG . Single-step methods for genomic evaluation in pigs , 2012,6(10):1565-1571. DOI:10.1017/S1751731112000742URL [本文引用: 1]
HENDERSON CR . Best linear unbiased estimation and prediction under a selection model , 1975,31(2):423-447. DOI:10.2307/2529430URL [本文引用: 1]
BALOCHEG, LEGARRAA, SALLEG, LARROQUEH, ASTRUC JM, ROBERT-GRANIEC, BARILLETF . Assessment of accuracy of genomic prediction for French Lacaune dairy sheep , 2014,97(2):1107-1116. DOI:10.3168/jds.2013-7135URL [本文引用: 1]
ZHONG SQ, DEKKERS JC, FERNANDO RL, JANNINK JL . Factors affecting accuracy from genomic selection in populations derived from multiple inbred lines: a barley case study , 2009,182(1):355-364. DOI:10.1534/genetics.108.098277URL [本文引用: 1]
MEUWISSEN TH . Accuracy of breeding values of ‘unrelated’ individuals predicted by dense SNP genotyping , 2009: 35-41. [本文引用: 2]
CALUS MP, MEUWISSEN TH ,DEROOS A P W, VEERKAMPR F. Accuracy of genomic selection using different methods to define haplotypes , 2008,178(1):553-561. DOI:10.1534/genetics.107.080838URL [本文引用: 1]
LUND MS ,ROOSA P D, VRIESA G D, DRUETT, DUCROCQV, FRITZS, GUILLAUMEF, GULDRANDSENB, LIUZ, REENTSR, SCHROOTENC, SEEFRIEDF, SUG S. A common reference population from four europeanholstein populations increases reliability of genomic predictions , 2011,43(1):43-51. DOI:10.1186/1297-9686-43-43 [本文引用: 2]
BEN JH, PHILLIP JB, AMANDA CC, VERBYLAK, GODDARD ME . Accuracy of genomic breeding values in multi-breed dairy cattle populations , 2009: 41-51. [本文引用: 1]
WEBER KL, THALLMAN RM, KEELE JW, SNELLING WM, BENNETT GL ,SMITHT P L, MCDANELDT G, ALLANM F, EENENNAAMA L V, KUEHNL A. Accuracy of genomic breeding values in multibreed beef cattle populations derived from deregressed breeding values and phenotypes , 2012,90(12):4177-4190. DOI:10.2527/jas.2011-4586URL [本文引用: 1]
GUOX, CHRISTENSEN OF, OSTERSENT, WANGY, LUND MS, SUG . Improving genetic evaluation of litter size and piglet mortality for both genotyped and nongenotyped individuals using a single-step method , 2015,93(2):503-512. DOI:10.2527/jas.2014-8331URL [本文引用: 1]
MIARY, PLASTOW GS, BRUCE HL , KEMPB . Genomic selection of pork pH in purebred pigs for crossbred performance 2014: VancouverBC,Canada. [本文引用: 1]
OUZ, TEMPELMAN RJ, STEIBEL JP, ERNST CW, BATES RO, BELLO NM . Genomic prediction accounting for residual heteroskedasticity ., 2015(6):1-13. [本文引用: 1]
ZHUB, WANG YH, NIUH, CHENY, ZHANG LP, GAO HJ, GAOX, LI JY, SUN SH . The Strategy of Parameter Optimization of Bayesian Methods for Genomic Selection in Livestock ScientiaAgriculturaSinica, 2014,47(22):4495-4505.(in Chinese) URL [本文引用: 1]
AGUILARI, MISZTALI, JOHNSON DL, LEGARRAA, TSURUTAS, LAWLOR TJ . Hot Topic: A unified approach to utilize phenotypic, full pedigree, and genomic information for genetic evaluation of Holstein final score1 , 2010,93(2):743-752. DOI:10.3168/jds.2009-2730URL [本文引用: 1]