Background: Different production systems and climates could lead to genotype-by-environment(G × E) interactions between populations, and the inclusion of G × E interactions is becoming essential in breeding ...Background: Different production systems and climates could lead to genotype-by-environment(G × E) interactions between populations, and the inclusion of G × E interactions is becoming essential in breeding decisions. The objective of this study was to investigate the performance of multi-trait models in genomic prediction in a limited number of environments with G × E interactions.Results: In total, 2,688 and 1,384 individuals with growth and reproduction phenotypes, respectively, from two Yorkshire pig populations with similar genetic backgrounds were genotyped with the PorcineSNP80 panel.Single-and multi-trait models with genomic best linear unbiased prediction(GBLUP) and BayesC π were implemented to investigate their genomic prediction abilities with 20 replicates of five-fold cross-validation.Our results regarding between-environment genetic correlations of growth and reproductive traits(ranging from 0.618 to 0.723) indicated the existence of G × E interactions between these two Yorkshire pig populations. For single-trait models, genomic prediction with GBLUP was only 1.1% more accurate on average in the combined population than in single populations, and no significant improvements were obtained by BayesC π for most traits. In addition, single-trait models with either GBLUP or BayesC π produced greater bias for the combined population than for single populations. However, multi-trait models with GBLUP and BayesC π better accommodated G × E interactions,yielding 2.2% – 3.8% and 1.0% – 2.5% higher prediction accuracies for growth and reproductive traits, respectively,compared to those for single-trait models of single populations and the combined population. The multi-trait models also yielded lower bias and larger gains in the case of a small reference population. The smaller improvement in prediction accuracy and larger bias obtained by the single-trait models in the combined population was mainly due to the low consistency of linkage disequilibrium between the two populations, which also caused the BayesC π method to always produce the largest standard error in marker effect estimation for the combined population.Conclusions: In conclusion, our findings confirmed that directly combining populations to enlarge the reference population is not efficient in improving the accuracy of genomic prediction in the presence of G × E interactions, while multi-trait models perform better in a limited number of environments with G × E interactions.展开更多
Yield prediction is the primary goal of genomic selection(GS)-assisted crop breeding.Because yield is a complex quantitative trait,making predictions from genotypic data is challenging.Transfer learning can produce an...Yield prediction is the primary goal of genomic selection(GS)-assisted crop breeding.Because yield is a complex quantitative trait,making predictions from genotypic data is challenging.Transfer learning can produce an effective model for a target task by leveraging knowledge from a different,but related,source domain and is considered a great potential method for improving yield prediction by integrating multi-trait data.However,it has not previously been applied to genotype-to-phenotype prediction owing to the lack of an efficient implementation framework.We therefore developed TrG2P,a transfer-learning-based framework.TrG2P first employs convolutional neural networks(CNN)to train models using non-yield-trait phenotypic and genotypic data,thus obtaining pre-trained models.Subsequently,the convolutional layer parameters from these pre-trained models are transferred to the yield prediction task,and the fully connected layers are retrained,thus obtaining fine-tuned models.Finally,the convolutional layer and the first fully connected layer of the fine-tuned models are fused,and the last fully connected layer is trained to enhance prediction performance.We applied TrG2P to five sets of genotypic and phenotypic data from maize(Zea mays),rice(Oryza sativa),and wheat(Triticum aestivum)and compared its model precision to that of seven other popular GS tools:ridge regression best linear unbiased prediction(rrBLUP),random forest,support vector regression,light gradient boosting machine(LightGBM),CNN,DeepGS,and deep neural network for genomic prediction(DNNGP).TrG2P improved the accuracy of yield prediction by 39.9%,6.8%,and 1.8%in rice,maize,and wheat,respectively,compared with predictions generated by the best-performing comparison model.Our work therefore demonstrates that transfer learning is an effective strategy for improving yield prediction by integrating information from non-yield-trait data.We attribute its enhanced prediction accuracy to the valuable information available from traits associated with yield and to training dataset augmentation.The Python implementation of TrG2P is available at https://github.com/lijinlong1991/TrG2P.The web-based tool is available at http://trg2p.ebreed.cn:81.展开更多
Two clonal trial stands of Chinese Fir （Cunninghamia lanceolata） were used in this study, one was 19-year-old stand which included 38 clones, and the other was 17-year-old stand including 102 clones.The statistical ...Two clonal trial stands of Chinese Fir （Cunninghamia lanceolata） were used in this study, one was 19-year-old stand which included 38 clones, and the other was 17-year-old stand including 102 clones.The statistical analyses showed that there were very significant genetic variations in height, DBH,volume and ratio of heartwood(R<sub>hw</sub>),wood basic density（ρ<sub>b</sub> ） of the clones in the two stands. The repeatability of clones was in median to high level,and the genetic CV was different over the all five traits.There were very significant phenotypic and genetic correlations among height,DBH and volume,and negative correlations among growth, R<sub>hw</sub> andρ<sub>b</sub>.The selection method experiment indicated that index selection could improve volume, R<sub>hw</sub> andρ<sub>b</sub>,showing synthetically superior selection effects compared to any individual trait selection methods.展开更多
Background Globally,the cultivation of cotton is constrained by its tendency for extended periods of growth.Early maturity plays a potential role in rainfed-based multiple cropping system especially in the current era...Background Globally,the cultivation of cotton is constrained by its tendency for extended periods of growth.Early maturity plays a potential role in rainfed-based multiple cropping system especially in the current era of climate change.In the current study,a set of 20 diverse Gossypium hirsutum genotypes were evaluated in two crop seasons with three planting densities and assessed for 11 morphological traits related to early maturity.The study aimed to identify genotype(s)that mature rapidly and accomplish well under diverse environmental conditions based on the two robust multivariate techniques called multi-trait stability index(MTSI)and multi-trait genotype-ideotype distance index(MGIDI).Results MTSI analysis revealed that out of the 20 genotypes,three genotypes,viz.,NNDC-30,A-2,and S-32 accomplished well in terms of early maturity traits in two seasons.Furthermore,three genotypes were selected using MGIDI method for each planting densities with a selection intensity of 15%.The strengths and weaknesses of the genotypes selected based on MGIDI method highlighted that the breeders could focus on developing early-maturing genotypes with specific traits such as days to first flower and boll opening.The selected genotypes exhibited positive genetic gains for traits related to earliness and a successful harvest during the first and second pickings.However,there were negative gains for traits related to flowering and boll opening.Conclusion The study identified three genotypes exhibiting early maturity and accomplished well under different planting densities.The multivariate methods(MTSI and MGIDI)serve as novel approaches for selecting desired genotypes in plant breeding programs,especially across various growing environments.These methods offer exclusive benefits and can easily construe and minimize multicollinearity issues.展开更多
本研究通过综合评价蚕豆品系产量性状在不同试点的丰产性、适应性和稳定性,筛选适应不同生态环境的产量性状稳定的优良品种(系)。同时评价各试点的区分力和代表性,为试点选择提供依据。2017年和2018年在甘肃和政县、康乐县、积石山县、...本研究通过综合评价蚕豆品系产量性状在不同试点的丰产性、适应性和稳定性,筛选适应不同生态环境的产量性状稳定的优良品种(系)。同时评价各试点的区分力和代表性,为试点选择提供依据。2017年和2018年在甘肃和政县、康乐县、积石山县、渭源县、临夏县和漳县6个试点分别种植5个蚕豆品系0215-1-4(L1)、0208-3-1(L2)、0208-3-2(L3)、0323-2-1(L4)、0161-1(L5)与1个对照品种和政尕蚕豆(L6),收获时记录株高、株粒数、小区产量、株荚数、分枝数、百粒重。采用联合方差和GGE(genotype+genotypes and environment interactions,GGE)双标图对产量性状进行基因型和基因型与环境互作分析。联合方差分析表明,6个农艺性状的基因型除小区产量和株高基因型与环境互作效应无显著差异外,其余性状的基因型与×环境互作效应均达到极显著水平(P<0.01);除株高和株粒数基因型×年份互作效应达到极显著水平外(P<0.01),其余农艺性状×年份互作效应无显著差异。相关性分析表明,小区产量与株荚数和株粒数正相关,与株荚数显著正相关(P<0.05),与百粒重负相关。GGE分析结果表明,品种(系)的适应性、丰产性和稳定性以及试点的区分力和代表性均具有较高的GGE变异值,变幅在78.54%~97.38%之间。蚕豆品系L3在康乐县、积石山县、渭源县和临夏县试点的产量适应性均较高,在和政县试点2018年产量适应性最高;丰产性高的品种(系)依次为L3>L2>L6>L4,稳定性最高的品种(系)依次为L4>L1>L5>L3。试点的区分力依次为康乐县2017年、积石山县2017年和2018年,试点的代表性依次为渭源县2017年、康乐县2018年、积石山县2018年。高产且稳定的品系是L3和L4,结合试点的区分力和代表性,最理想的生态区试点是积石山县。本研究利用GGE双标图对甘肃蚕豆参试品种进行产量组分性状分析,为蚕豆品种综合评价提供参考。展开更多
The objectives of this study were to estimate genetic parameters of lactation average somatic cell scores （LSCS） and examine genetic associations between LSCS and production traits in the first three lactations of C...The objectives of this study were to estimate genetic parameters of lactation average somatic cell scores （LSCS） and examine genetic associations between LSCS and production traits in the first three lactations of Chinese Holstein cows using single-parity multi-trait animal model and multi-trait repeatability animal model. There were totally 273605 lactation records of Chinese Holstein cows with first calving from 2001 to 2012. Heritability estimates for LSCS ranged from 0.144 to 0.187. Genetic correlations between LSCS and 305 days milk, protein percentage and fat percentage were -0.079, -0.082 and -0.135, respectively. Phenotypic correlation between LSCS and 305 days milk yield was negative （-0.103 to -0.190）. Genetic correlation between 305 days milk and fat percentage or protein percentage was highly negative. Genetic correlation between milk fat percentage and milk protein percentage was highly favorable. Heritabilities of production traits decreased with increase of parity, whereas heritability of LSCS increased with increase of parity.展开更多
基金supported by grants from the earmarked fund for China Agriculture Research System (CARS-35)Modern Agriculture Science and Technology Key Project of Hebei Province (19226376D)+2 种基金the National Key Research and Development Project (SQ2019YFE00771)the National Natural Science Foundation of China (31671327)Major Project of Selection for New Livestock and Poultry Breeds of Zhejiang Province (2016C02054–5)。
文摘Background: Different production systems and climates could lead to genotype-by-environment(G × E) interactions between populations, and the inclusion of G × E interactions is becoming essential in breeding decisions. The objective of this study was to investigate the performance of multi-trait models in genomic prediction in a limited number of environments with G × E interactions.Results: In total, 2,688 and 1,384 individuals with growth and reproduction phenotypes, respectively, from two Yorkshire pig populations with similar genetic backgrounds were genotyped with the PorcineSNP80 panel.Single-and multi-trait models with genomic best linear unbiased prediction(GBLUP) and BayesC π were implemented to investigate their genomic prediction abilities with 20 replicates of five-fold cross-validation.Our results regarding between-environment genetic correlations of growth and reproductive traits(ranging from 0.618 to 0.723) indicated the existence of G × E interactions between these two Yorkshire pig populations. For single-trait models, genomic prediction with GBLUP was only 1.1% more accurate on average in the combined population than in single populations, and no significant improvements were obtained by BayesC π for most traits. In addition, single-trait models with either GBLUP or BayesC π produced greater bias for the combined population than for single populations. However, multi-trait models with GBLUP and BayesC π better accommodated G × E interactions,yielding 2.2% – 3.8% and 1.0% – 2.5% higher prediction accuracies for growth and reproductive traits, respectively,compared to those for single-trait models of single populations and the combined population. The multi-trait models also yielded lower bias and larger gains in the case of a small reference population. The smaller improvement in prediction accuracy and larger bias obtained by the single-trait models in the combined population was mainly due to the low consistency of linkage disequilibrium between the two populations, which also caused the BayesC π method to always produce the largest standard error in marker effect estimation for the combined population.Conclusions: In conclusion, our findings confirmed that directly combining populations to enlarge the reference population is not efficient in improving the accuracy of genomic prediction in the presence of G × E interactions, while multi-trait models perform better in a limited number of environments with G × E interactions.
基金This research was funded by the STI2030-Major Projects(no.2023ZD0406104)the Beijing Postdoctoral Research Foundation(no.2023-ZZ-116).
文摘Yield prediction is the primary goal of genomic selection(GS)-assisted crop breeding.Because yield is a complex quantitative trait,making predictions from genotypic data is challenging.Transfer learning can produce an effective model for a target task by leveraging knowledge from a different,but related,source domain and is considered a great potential method for improving yield prediction by integrating multi-trait data.However,it has not previously been applied to genotype-to-phenotype prediction owing to the lack of an efficient implementation framework.We therefore developed TrG2P,a transfer-learning-based framework.TrG2P first employs convolutional neural networks(CNN)to train models using non-yield-trait phenotypic and genotypic data,thus obtaining pre-trained models.Subsequently,the convolutional layer parameters from these pre-trained models are transferred to the yield prediction task,and the fully connected layers are retrained,thus obtaining fine-tuned models.Finally,the convolutional layer and the first fully connected layer of the fine-tuned models are fused,and the last fully connected layer is trained to enhance prediction performance.We applied TrG2P to five sets of genotypic and phenotypic data from maize(Zea mays),rice(Oryza sativa),and wheat(Triticum aestivum)and compared its model precision to that of seven other popular GS tools:ridge regression best linear unbiased prediction(rrBLUP),random forest,support vector regression,light gradient boosting machine(LightGBM),CNN,DeepGS,and deep neural network for genomic prediction(DNNGP).TrG2P improved the accuracy of yield prediction by 39.9%,6.8%,and 1.8%in rice,maize,and wheat,respectively,compared with predictions generated by the best-performing comparison model.Our work therefore demonstrates that transfer learning is an effective strategy for improving yield prediction by integrating information from non-yield-trait data.We attribute its enhanced prediction accuracy to the valuable information available from traits associated with yield and to training dataset augmentation.The Python implementation of TrG2P is available at https://github.com/lijinlong1991/TrG2P.The web-based tool is available at http://trg2p.ebreed.cn:81.
文摘Two clonal trial stands of Chinese Fir （Cunninghamia lanceolata） were used in this study, one was 19-year-old stand which included 38 clones, and the other was 17-year-old stand including 102 clones.The statistical analyses showed that there were very significant genetic variations in height, DBH,volume and ratio of heartwood(R<sub>hw</sub>),wood basic density（ρ<sub>b</sub> ） of the clones in the two stands. The repeatability of clones was in median to high level,and the genetic CV was different over the all five traits.There were very significant phenotypic and genetic correlations among height,DBH and volume,and negative correlations among growth, R<sub>hw</sub> andρ<sub>b</sub>.The selection method experiment indicated that index selection could improve volume, R<sub>hw</sub> andρ<sub>b</sub>,showing synthetically superior selection effects compared to any individual trait selection methods.
文摘Background Globally,the cultivation of cotton is constrained by its tendency for extended periods of growth.Early maturity plays a potential role in rainfed-based multiple cropping system especially in the current era of climate change.In the current study,a set of 20 diverse Gossypium hirsutum genotypes were evaluated in two crop seasons with three planting densities and assessed for 11 morphological traits related to early maturity.The study aimed to identify genotype(s)that mature rapidly and accomplish well under diverse environmental conditions based on the two robust multivariate techniques called multi-trait stability index(MTSI)and multi-trait genotype-ideotype distance index(MGIDI).Results MTSI analysis revealed that out of the 20 genotypes,three genotypes,viz.,NNDC-30,A-2,and S-32 accomplished well in terms of early maturity traits in two seasons.Furthermore,three genotypes were selected using MGIDI method for each planting densities with a selection intensity of 15%.The strengths and weaknesses of the genotypes selected based on MGIDI method highlighted that the breeders could focus on developing early-maturing genotypes with specific traits such as days to first flower and boll opening.The selected genotypes exhibited positive genetic gains for traits related to earliness and a successful harvest during the first and second pickings.However,there were negative gains for traits related to flowering and boll opening.Conclusion The study identified three genotypes exhibiting early maturity and accomplished well under different planting densities.The multivariate methods(MTSI and MGIDI)serve as novel approaches for selecting desired genotypes in plant breeding programs,especially across various growing environments.These methods offer exclusive benefits and can easily construe and minimize multicollinearity issues.
文摘本研究通过综合评价蚕豆品系产量性状在不同试点的丰产性、适应性和稳定性,筛选适应不同生态环境的产量性状稳定的优良品种(系)。同时评价各试点的区分力和代表性,为试点选择提供依据。2017年和2018年在甘肃和政县、康乐县、积石山县、渭源县、临夏县和漳县6个试点分别种植5个蚕豆品系0215-1-4(L1)、0208-3-1(L2)、0208-3-2(L3)、0323-2-1(L4)、0161-1(L5)与1个对照品种和政尕蚕豆(L6),收获时记录株高、株粒数、小区产量、株荚数、分枝数、百粒重。采用联合方差和GGE(genotype+genotypes and environment interactions,GGE)双标图对产量性状进行基因型和基因型与环境互作分析。联合方差分析表明,6个农艺性状的基因型除小区产量和株高基因型与环境互作效应无显著差异外,其余性状的基因型与×环境互作效应均达到极显著水平(P<0.01);除株高和株粒数基因型×年份互作效应达到极显著水平外(P<0.01),其余农艺性状×年份互作效应无显著差异。相关性分析表明,小区产量与株荚数和株粒数正相关,与株荚数显著正相关(P<0.05),与百粒重负相关。GGE分析结果表明,品种(系)的适应性、丰产性和稳定性以及试点的区分力和代表性均具有较高的GGE变异值,变幅在78.54%~97.38%之间。蚕豆品系L3在康乐县、积石山县、渭源县和临夏县试点的产量适应性均较高,在和政县试点2018年产量适应性最高;丰产性高的品种(系)依次为L3>L2>L6>L4,稳定性最高的品种(系)依次为L4>L1>L5>L3。试点的区分力依次为康乐县2017年、积石山县2017年和2018年,试点的代表性依次为渭源县2017年、康乐县2018年、积石山县2018年。高产且稳定的品系是L3和L4,结合试点的区分力和代表性,最理想的生态区试点是积石山县。本研究利用GGE双标图对甘肃蚕豆参试品种进行产量组分性状分析,为蚕豆品种综合评价提供参考。
基金fundings from the National Natural Science Foundation of China (31200927)the National Modern Agricultural Industry Technology Fund for Scientists in Sheep Industry System, China (CARS-39-04B)+1 种基金the National Key Technology Research and Development Program of the Ministry of Science and Technology of China (2011BAD28B02, 2012BAD12B06)the Chinese Academy of Agricultural Sciences Foundation (2012cj-2)
文摘The objectives of this study were to estimate genetic parameters of lactation average somatic cell scores （LSCS） and examine genetic associations between LSCS and production traits in the first three lactations of Chinese Holstein cows using single-parity multi-trait animal model and multi-trait repeatability animal model. There were totally 273605 lactation records of Chinese Holstein cows with first calving from 2001 to 2012. Heritability estimates for LSCS ranged from 0.144 to 0.187. Genetic correlations between LSCS and 305 days milk, protein percentage and fat percentage were -0.079, -0.082 and -0.135, respectively. Phenotypic correlation between LSCS and 305 days milk yield was negative （-0.103 to -0.190）. Genetic correlation between 305 days milk and fat percentage or protein percentage was highly negative. Genetic correlation between milk fat percentage and milk protein percentage was highly favorable. Heritabilities of production traits decreased with increase of parity, whereas heritability of LSCS increased with increase of parity.
