基于全基因组的等速蹬踏伸肌力训练效果预测模型

Prediction Model for Resistance Training Effects of Isokinetic Knee Extensor Strength Based on GWAS

目的:基于全基因组关联分析(GWAS)筛选抗阻训练后等速蹬踏伸肌力训练效果相关的遗传标记,分析遗传标记的可能生物学作用,并建立训练效果预测模型,为个性化健身指导方案的制定提供参考。方法:193名健康成年人完成12周抗阻训练,干预前后测试表型指标(等速蹬踏伸肌力、肌肉厚度、肌肉质量)。采集静脉血并提取DNA,通过Infinium Chinese Genotyping Array v1.0芯片进行全基因组基因分型,PLINK1.9进行全基因组关联分析。采用Lasso回归降维变量,采用Logistic回归、Nomogram列线图和逐步线性回归建立等速蹬踏伸肌力训练效果预测模型。采用HaploReg v4.2、SNPnexus工具对筛选出的遗传标记进行生物信息学分析。结果:1)12周抗阻训练干预后,受试者等速蹬踏伸肌力整体显著增加(Δ=18.02%,p<0.01,ES=0.32),变化范围-58.48%~109.45%。2)rs1419957、rs12326802、rs8010482等19个SNPs与12周抗阻训练后等速蹬踏伸肌力变化百分比显著关联(p<1×10^(-5))。3)以等速蹬踏伸肌力、右侧股直肌和股中间肌厚度初始值、rs35199905、rs12326802、144298896建立的Logistic模型曲线下面积AUC=0.896;以等速蹬踏伸肌力、躯干肌肉含量、rs8010482、rs144298896、rs12326802、rs35199905、rs17116169、rs61973994、rs1419957建立的逐步回归预测模型R2=0.67。4)生物信息学分析显示,SNPs可调控基因模体,7个SNPs可能在DNA修复、染色质结构、基因表达等REACTOME pathway上发挥作用。结论:等速蹬踏伸肌力训练效果存在个体差异,rs1419957、rs12326802、rs8010482等19个SNPs是训练敏感的SNPs,这些SNPs可能在DNA修复、染色质结构、基因表达等REACTOME pathway上发挥作用。以遗传标记和表型指标建立的等速蹬踏伸肌力训练效果的Logistic预测模型、Nomogram列线图和逐步回归模型可用于等速蹬踏伸肌力训练效果预测,这为个性化健身指导方案的制定提供了参考。

Purpose:This study aimed to screen genetic markers associated with the training effects of isokinetic knee extensor strength after resistance training based on genome-wide association study(GWAS),analyze the potential biological roles of these genetic markers,and establish a predictive model for training effects to provide reference for personalized fitness guidance.Methods:193 healthy adults completed 12 weeks of resistance training,and phenotypic indicators(isokinetic knee extensor strength,muscle thickness,muscle quality) were measured before and after the intervention.Venous blood was collected,and DNA was extracted for whole-genome genotyping using the Infinium Chinese Genotyping Array v1.0 chip,and whole-genome association analysis was performed using PLINK1.9.Lasso regression was used for variable selection,and logistic regression,Nomogram column line graph,and stepwise linear regression were employed to establish a predictive model for isokinetic knee extensor strength training effects.The selected genetic markers were subjected to bioinformatics analysis using HaploReg v4.2and SNPnexus tools.Results:1) After 12 weeks of strength training intervention,participants showed a significant overall increase in isokinetic knee extensor strength(Δ=18.02%,p<0.01,ES=0.32),with individual differences ranging from -58.48% to 109.45%.2) 19 SNPs,including rs1419957,rs12326802,and rs8010482,were significantly associated with the percentage change in isokinetic knee extensor strength after 12 weeks of resistance training(p<1×10^(-5)).3) Logistic regression models based on initial values of isokinetic knee extensor strength,right vastus medialis and intermedius thickness,rs35199905,rs12326802,and 144298896,had an area under the curve(AUC) of 0.896;Stepwise regression models based on isokinetic knee extensor strength,trunk muscle content,rs8010482,rs144298896,rs12326802,rs35199905,rs17116169,rs61973994,and rs1419957 had an R2of 0.67.4) Bioinformatics analysis revealed that SNPs can regulate gene motifs,and 7 SNPs may be involved in DNA repair,chromatin structure,gene expression,and other REACTOME pathways.Conclusion:Individual differences exist in the training effects of isokinetic knee extensor strength,and 19 SNPs,including rs1419957,rs12326802,and rs8010482,are sensitive to training effects,with potential roles in DNA repair,chromatin structure,gene expression,and other REACTOME pathways.The logistic predictive model,Nomogram column line graph,and stepwise regression model based on genetic markers and phenotypic indicators can be used to predict the training effects of isokinetic knee extensor strength,providing valuable references for personalized fitness guidance.