基于机器学习的猪生长性状基因组预测

doi:10.16288/j.yczz.23-120

遗传 ›› 2023, Vol. 45 ›› Issue (10): 922-932.doi: 10.16288/j.yczz.23-120

基于机器学习的猪生长性状基因组预测

陈栋^1,^2,³(), 王书杰^1,^2,³, 赵真坚^1,^2,³, 姬祥^1,^2,³, 申琦^1,^2,³, 余杨^1,^2,³, 崔晟頔^1,^2,³, 王俊戈^1,^2,³, 陈子旸^1,^2,³, 王金勇⁴, 郭宗义⁴, 吴平先⁴, 唐国庆^1,^2,³()

1.四川农业大学动物科技学院，农业农村部畜禽生物组学重点实验室，成都 611130
2.四川农业大学，畜禽遗传资源发掘与创新利用四川省重点实验室，成都 611130
3.四川农业大学动物科技学院，猪禽种业全国重点实验室，成都 611130
4.国家生猪技术创新中心，重庆 402460

收稿日期:2023-04-26 修回日期:2023-08-14 出版日期:2023-10-20 发布日期:2023-08-16
通讯作者: 唐国庆 E-mail:1123278154@qq.com;tyq003@163.com
作者简介:陈栋，硕士，专业方向：畜牧学。E-mail: 1123278154@qq.com
基金资助:
国家生猪技术创新中心先导科技项目(NCTIP-XD/B01);四川省科技厅项目(2020YFN0024);四川省科技厅项目(2021ZDZX0008);四川省科技厅项目(2021YFYZ0030);四川省猪创新团队项目(sccxtd-2022-08)

Genomic prediction of pig growth traits based on machine learning

Dong Chen^1,^2,³(), Shujie Wang^1,^2,³, Zhenjian Zhao^1,^2,³, Xiang Ji^1,^2,³, Qi Shen^1,^2,³, Yang Yu^1,^2,³, Shengdi Cui^1,^2,³, Junge Wang^1,^2,³, Ziyang Chen^1,^2,³, Jinyong Wang⁴, Zongyi Guo⁴, Pingxian Wu⁴, Guoqing Tang^1,^2,³()

1. Key Laboratory of Livestock and Poultry Multi-omics, Ministry of Agriculture and Rural Affairs, College of Animal Science and Technology, Sichuan Agricultural University, Chengdu 611130, China
2. Farm Animal Genetic Resources Exploration and Innovation Key Laboratory of Sichuan Province, Sichuan Agricultural University, Chengdu 611130, China
3. State Key Laboratory of Swine and Poultry Breeding Industry, College of Animal Science and Technology, Sichuan Agricultural University, Chengdu 611130, China
4. National Center of Technology Innovation for Pigs, Chongqing 402460, China

Received:2023-04-26 Revised:2023-08-14 Published:2023-10-20 Online:2023-08-16
Contact: Guoqing Tang E-mail:1123278154@qq.com;tyq003@163.com
Supported by:
Strategic Priority Research Program of the National Center of Technology Innovation for Pigs(NCTIP-XD/B01);Sichuan Science and Technology Program(2020YFN0024);Sichuan Science and Technology Program(2021ZDZX0008);Sichuan Science and Technology Program(2021YFYZ0030);Sichuan Innovation Team of Pig(sccxtd-2022-08)

摘要/Abstract

摘要：

为了比较自动机器学习下不同机器学习模型预测部分猪生长性状与全基因组估计育种值(genomic estimated breeding value，GEBV)的性能，并寻找适合的机器学习模型，以优化生猪育种的全基因组评估方法，本研究利用来自多个公司9968头猪的基因组信息、系谱矩阵、固定效应及表型信息通过自动机器学习方法获取深度学习(deep learning，DL)、随机森林(random forest，RF)、梯度提升机(gradient boosting machine，GBM)和极致梯度提升(extreme gradient boosting，XGB)4种机器学习最佳模型。采用10折交叉验证分别对猪达100 kg校正背膘(correcting backfat to 100 kg，B100)、达115 kg校正背膘(correcting backfat to 115 kg，B115)、达100 kg校正日龄(correcting days to 100 kg，D100)、达115 kg校正日龄(correcting days to 100 kg，D115)的GEBV及其表型进行预测，比较不同机器学习模型应用于猪基因组评估的性能。结果表明：机器学习模型对GEBV的估计准确性高于性状表型；在GEBV预测中，GBM在B100、B115、D100、D115的预测准确性分别为0.683、0.710、0.866、0.871，略高于其他方法；在表型预测中，对猪B100、B115、D100、D115预测性能最好的模型依次为GBM(0.547)、DL(0.547)、XGB(0.672、0.670)；在模型训练所需时间上，RF远高于其他3种模型，GBM与DL居中，XGB所需时间最少。综上所述，通过自动机器学习获取的机器学习模型对GEBV预测的准确性高于表型；GBM模型总体上表现出最高的预测准确性与较短训练时间；XGB能够利用最短的时间训练准确性较高的预测模型；RF模型的训练时间远超其他3种模型，且准确性不足，不适用猪生长性状表型与GEBV预测。

关键词: 基因组估计育种值, 生长性状, 自动机器学习, 性能比较

Abstract:

This study aimed to assess and compare the performance of different machine learning models in predicting selected pig growth traits and genomic estimated breeding values (GEBV) using automated machine learning, with the goal of optimizing whole-genome evaluation methods in pig breeding. The research employed genomic information, pedigree matrices, fixed effects, and phenotype data from 9968 pigs across multiple companies to derive four optimal machine learning models: deep learning (DL), random forest (RF), gradient boosting machine (GBM), and extreme gradient boosting (XGB). Through 10-fold cross-validation, predictions were made for GEBV and phenotypes of pigs reaching weight milestones (100 kg and 115 kg) with adjustments for backfat and days to weight. The findings indicated that machine learning models exhibited higher accuracy in predicting GEBV compared to phenotypic traits. Notably, GBM demonstrated superior GEBV prediction accuracy, with values of 0.683, 0.710, 0.866, and 0.871 for B100, B115, D100, and D115, respectively, slightly outperforming other methods. In phenotype prediction, GBM emerged as the best-performing model for pigs with B100, B115, D100, and D115 traits, achieving prediction accuracies of 0.547, followed by DL at 0.547, and then XGB with accuracies of 0.672 and 0.670. In terms of model training time, RF required the most time, while GBM and DL fell in between, and XGB demonstrated the shortest training time. In summary, machine learning models obtained through automated techniques exhibited higher GEBV prediction accuracy compared to phenotypic traits. GBM emerged as the overall top performer in terms of prediction accuracy and training time efficiency, while XGB demonstrated the ability to train accurate prediction models within a short timeframe. RF, on the other hand, had longer training times and insufficient accuracy, rendering it unsuitable for predicting pig growth traits and GEBV.

Key words: genomic estimated breeding values, growth traits, automated machine learning, performance comparison

陈栋, 王书杰, 赵真坚, 姬祥, 申琦, 余杨, 崔晟頔, 王俊戈, 陈子旸, 王金勇, 郭宗义, 吴平先, 唐国庆. 基于机器学习的猪生长性状基因组预测[J]. 遗传, 2023, 45(10): 922-932.

Dong Chen, Shujie Wang, Zhenjian Zhao, Xiang Ji, Qi Shen, Yang Yu, Shengdi Cui, Junge Wang, Ziyang Chen, Jinyong Wang, Zongyi Guo, Pingxian Wu, Guoqing Tang. Genomic prediction of pig growth traits based on machine learning[J]. Hereditas(Beijing), 2023, 45(10): 922-932.

图/表 11

表1

表2

表3

表4

表5

图1

图2

图3

图4

图5

图6

参考文献 30

[1]	Wang KJ, Liu YF, Xu Q, Liu CK, Wang J, Ding C, Fang MY. A post-GWAS confirming GPAT3 gene associated with pig growth and a significant SNP influencing its promoter activity. Anim Genet, 2017, 48(4): 478-482.
[2]	Guo YM, Huang YX, Hou LJ, Ma JW, Chen CY, Ai HS, Huang LS, Ren J. Genome-wide detection of genetic markers associated with growth and fatness in four pig populations using four approaches. Genet Sel Evol, 2017, 49(1): 21.
[3]	Ding RR, Yang M, Wang XW, Quan JP, Zhuang ZW, Zhou SP, Li SY, Xu Z, Zheng EQ, Cai GY, Liu DW, Huang W, Yang J, Wu ZF. Genetic architecture of feeding behavior and feed efficiency in a Duroc pig population. Front Genet, 2018, 9: 220.
[4]	Vanraden PM. Symposium review:How to implement genomic selection. J Dairy Sci, 2020, 103(6): 5291-5301.
[5]	Meuwissen THE, Hayes BJ, Goddard ME. Prediction of total genetic value using genome-wide dense marker maps. Genetics, 2001, 157(4): 1819-1829.
[6]	Guo P, Zhang JB, Cao S. Study on Bayesian method and machine learning genome-wide selection of milk production traits in dairy cows. Heilongjiang Anim Sci Vet Med, 2023(5): 56-60+64.
	郭鹏, 张建斌, 曹晟. 奶牛产奶性状贝叶斯方法与机器学习全基因组选择研究. 黑龙江畜牧兽医, 2023(5): 56-60+64.
[7]	Akanno EC, Schenkel FS, Sargolzaei M, Friendship RM, Robinson JAB. Opportunities for genome-wide selection for pig breeding in developing countries. J Anim Sci, 2013, 91(10): 4617-4627.
[8]	Samorè AB, Fontanesi L. Genomic selection in pigs: State of the art and perspectives. Ital J Anim Sci, 2016, 15: 211-232.
[9]	Marchini J, Howie B. Genotype imputation for genome- wide association studies. Nat Rev Genet, 2010, 11(7): 499-511.
[10]	Das S, Forer L, Schönherr S, Sidore C, Locke AE, Kwong A, Vrieze SI, Chew EY, Levy S, Mcgue M, Schlessinger D, Stambolian D, Loh PR, Iacono WG, Swaroop A, Scott LJ, Cucca F, Kronenberg F, Boehnke M, Abecasis GR, Fuchsberger C. Next-generation genotype imputation service and methods. Nat Genet, 2016, 48(10): 1284-1287.
[11]	Sollero BP, Howard JT, Spangler ML.The impact of reducing the frequency of animals genotyped at higher density on imputation and prediction accuracies using ssGBLUP1. J Anim Sci, 2019, 97(7): 2780-2792.
[12]	Fernandes Júnior GA, Carvalheiro R, de Oliveira HN, Sargolzaei M, Costilla R, Ventura RV, Fonseca LFS, Neves HHR, Hayes BJ, Imputation accuracy to whole-genome sequence in Nellore cattle. Genet Sel Evol, 2021, 53(1): 27.
[13]	Alves AAC, Espigolan R, Bresolin T, Costa RM, Fernandes Júnior GA, Ventura RV, Carvalheiro R, Albuquerque LG. Genome-enabled prediction of reproductive traits in Nellore cattle using parametric models and machine learning methods. Anim Genet, 2021, 52(1): 32-46.
[14]	Zhao W, Lai XS, Liu DY, Zhang ZY, Ma PP, Wang QS, Zhang Z, Pan YC. Applications of support vector machine in genomic prediction in pig and maize populations. Front Genet, 2020, 11: 598318.
[15]	Montesinos López OA, Montesinos López A, Crossa J.Random Forest for Genomic Prediction. In: Multivariate Statistical Machine Learning Methods for Genomic Prediction. Cham: Springer International Publishing, 2022, 633-681.
[16]	Liang M.Research on genome-wide selection based on machine learning algorithm[Dissertation]. Chinese Academy of Agricultural Sciences, 2021.
	梁忙.基于机器学习算法的全基因组选择研究[学位论文]. 中国农业科学院, 2021.
[17]	Barreiro E, Munteanu CR, Cruz-Monteagudo M, Pazos A, González-Díaz H. Net-Net Auto Machine Learning (AutoML) Prediction of Complex Ecosystems. Scientific reports. 2018; 8(1): 12340.
[18]	Romero RAA, Deypalan MNY, Mehrotra S, Jungao JT, Sheils NE, Manduchi E, Moore JH. Benchmarking AutoML frameworks for disease prediction using medical claims. BioData Min, 2022, 15(1): 15.
[19]	Friedman JH. Greedy function approximation: A gradient boosting machine. Ann Statist, 2001, 29(5): 1189-1232.
[20]	Natekin A, Knoll A. Gradient boosting machines, a tutorial. Front Neurorobot, 2013, 7: 21.
[21]	Abdollahi-Arpanahi R, Gianola D, Peñagaricano F. Deep learning versus parametric and ensemble methods for genomic prediction of complex phenotypes. Genet Sel Evol, 2020, 52(1): 12.
[22]	Breiman L. Random Forests. Machine Learning, 2001, 45: 5-32.
[23]	González-recio O, Forni S. Genome-wide prediction of discrete traits using Bayesian regressions and machine learning. Genet Sel Evol, 2011, 43(1): 7.
[24]	Wang FY, Wang YC, Ji XK, Wang ZP. Effective macrosomia prediction using random forest algorithm. Int J Environ Res Public Health, 2022, 19(6): 3245.
[25]	Chen TQ, Guestrin C. XGBoost: a scalable tree boosting system. In:Proceedings of the 22nd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining, 2016, 785-794.
[26]	Hou NZ, Li MZ, He L, Xie B, Wang L, Zhang RM, Yu Y, Sun XD, Pan ZS, Wang K. Predicting 30-days mortality for MIMIC-III patients with sepsis-3: a machine learning approach using XGboost. J Transl Med, 2020, 18(1): 462.
[27]	Tan XF, Li GS. Overview of deep learning development. In: The 15th National Security Geophysics Symposium. 2019.
	谭笑枫, 李广帅.深度学习发展综述. 见: 第十五届国家安全地球物理专题研讨会. 2019.
[28]	Nayeri S, Sargolzaei M, Tulpan D. A review of traditional and machine learning methods applied to animal breeding. Anim Health Res Rev, 2019, 20(1): 31-46.
[29]	Zheye Peng, Zijun Tang, Minzhu Xie. Research progress in machine learning methods for gene-gene interaction detection. Hereditas(Beijing), 2018, 40(3): 218-226.
	彭哲也, 唐紫珺, 谢民主. 机器学习方法在基因交互作用探测中的研究进展. 遗传, 2018, 40(3): 218-226.
[30]	Li B, Zhang NX, Wang YG, George AW, Reverter A, Li YT. Genomic prediction of breeding values using a subset of snps identified by three machine learning methods. Front Genet, 2018, 9: 237.

编辑推荐

Metrics

www.chinagene.cn
备案号：京ICP备09063187号-4
总访问:,今日访问:,当前在线:

公司	芯片、系谱及固定效应数据	校正背膘表型数据	校正日龄表型数据
巨星农牧	2008	1554	1554
新希望	2093	2002	1997
明兴	1355	1342	1342
铁骑力士	2083	859	859
天兆	1575	1261	508
德康	854	853	853
合计	9968	7871	7113

参数名	GEBV估计参数	表型预测参数
ntrees	61	35
max_depth	7	8
min_rows	10	100
distribution	gaussian	gaussian
sample_rate	0.8	0.6
col_sample_rate	0.8	1
col_sample_rate_per_tree	0.8	1
histogram_type	UniformAdaptive	UniformAdaptive
categorical_encoding	Enum	Enum

参数名	GEBV估计参数	表型预测参数
ntrees	11	15
max_depth	20	20
sample_rate	0.632	0.632
histogram_type	UniformAdaptive	UniformAdaptive
categorical_encoding	Enum	Enum
distribution	gaussian	gaussian

参数名	GEBV估计参数	表型预测参数
distribution	gaussian	gaussian
categorical_encoding	OneHotInternal	OneHotInternal
ntrees	111	71
max_depth	5	10
learn_rate	0.3	5
col_sample_rate	0.8	0.8
colsample_bylevel	0.8	0.8
tree_method	exact	exact
booster	gbtree	dart
reg_lambda	0.001	100
reg_alpha	0.1	1
dmatrix_type	sparse	sparse
backend	cpu	cpu

参数	GEBV估计参数	表型预测参数
activation	RectifierWithDropout	RectifierWithDropout
hidden	[100,100,100]	[100, 100, 100]
epochs	222	315
epsilon	1.00E-09	1.00E-08
rate	0.005	0.005
input_dropout_ratio	0.2	0.05
hidden_dropout_ratios	[0.2, 0.2, 0.2]	[0.2, 0.2, 0.2]
distribution	gaussian	gaussian
tweedie_power	1.5	1.5
categorical_encoding	OneHotInternal	OneHotInternal

基于机器学习的猪生长性状基因组预测

Genomic prediction of pig growth traits based on machine learning

RichHTML

PDF (PC)

可视化

摘要/Abstract

引用本文

使用本文

图/表 11

参考文献 30

相关文章 12

编辑推荐

Metrics

[1]	王春玲孟晨玲曹少先张俊孟春花王慧利方永飞朱冬冬茆达干. 湖羊、东湖杂种羊POMC基因外显子3单核苷酸多态性及其与生长性状的关联分析[J]. 遗传, 2013, 35(9): 1095-1100.
[2]	王鹏宇，官却扎西，祁全青，德毛，张文广，李金泉. 母体遗传效应对青海细毛羊生产性能遗传参数估计的影响[J]. 遗传, 2012, 34(5): 584-590.
[3]	朱媛媛，梁宏伟，李忠，罗相忠，李林，张志伟，邹桂伟. 黄颡鱼MSTN基因多态性及其与生长性状的相关性分析[J]. 遗传, 2012, 34(1): 72-78.
[4]	杜智恒，刘宗岳，白秀娟. 北极狐GHR基因单核苷酸多态性及其与生长性状的相关性分析[J]. 遗传, 2010, 32(6): 599-605.
[5]	樊佳佳，白俊杰，李小慧，何小平，何小燕，李胜杰，叶星，吴立新 . 大口黑鲈生长性状的微卫星DNA标记筛选[J]. 遗传, 2009, 31(5): 515-515―522.
[6]	张春雷，王艳红，陈宏，雷初朝，房兴堂，王居强，马桂变，牛晖，肖杰. 牛POMC基因多态性及其与南阳牛生长性状的相关分析[J]. 遗传, 2009, 31(12): 1221-1225.
[7]	陶勇，李国辉，胡玉萍，M Dafalla Mekki，陈宽维，王金玉. MC4R、POU1F1基因对京海黄鸡生长性能的遗传效应[J]. 遗传, 2008, 30(7): 900-906.
[8]	李世鹏，杜智恒，宁方勇，孙洪霞，白秀娟. 鸽黑素皮质素受体3、4(MC3R、MC4R)基因多态性与生长性状的相关分析[J]. 遗传, 2008, 30(10): 1333-1340.
[9]	张争锋，陈宏，李秋玲，雷初朝，张春雷，王新庄，王居强，王轶敏. 南阳牛DGAT2基因PCR-RFLP多态性及其与生长性状相关性研究[J]. 遗传, 2007, 29(8): 945-945―950.
[10]	赵晓枫，徐宁迎，胡晓湘，李宁. IGF1调控区微卫星座位对金华猪生长性能的影响[J]. 遗传, 2007, 29(2): 206-206―210.
[11]	高雪，徐秀容，任红艳，张英汉，许尚忠. 不同生长时期基因调控对南阳牛生长发育的影响[J]. 遗传, 2006, 28(8): 927-932.
[12]	王青来，王爱国，吴珍芳，张豪，罗旭芳. 应用VCE4.0估计长白猪生长性状的遗传参数[J]. 遗传, 2004, 26(6): 811-814.