Prediction Model of Bench Press 1RM Resistance Training Effect in Young People Based on GWAS

Objective: The genome-wide association analysis(GWAS) was used to screen genetic markers associated with resistance training effects of bench press 1RM, then a predictive model of bench press 1RM resistance training effect was constructed by combining with phenotye indicators; in addition, the bioinformatics was used to analyze the possible mechanisms of genetic markers in order to provide references for formulate precise exercise and fitness programs. Methods: 193 non-regular exercise adults completed a 12-week resistance training program, the phenotypic indicators such as bench press 1RM, body composition, and muscle thickness were tested before and after the training intervention. The subjects' DNA was collected, and whole-genome genotyping analysis was performed using the Illumina Infinium CGA-24v1-0 chip. GWAS analysis was conducted using PLINK1.9software, and single-nucleotide polymorphisms(SNPs) affecting bench press 1RM training effect were screened. The mean method was used to calculate the genomics polygenic scores(GPGS) after weighting. The K-mean method was used to cluster the percentage change in bench press 1RM and the forward logistic regression analysis was used to build a comprehensive model for predicting the training effect of genomic-phenotye indicators. HaploReg v4.1, GTEx, and KEGG were used to annotate the SNPs included in the model for biofunctionality. Results: 1) After 12 weeks of resistance training, the mean improvement in bench press 1RM was 36.25%(P<0.01), but the individual variation was significant which was ranging from-31.25% to 176.92%. 2) GWAS showed 35 SNPs were significantly associated with the effect of bench press 1RM resistance training(P<1×10-6), and 7 SNPs were included in the genomics regression model to explain 39.6% of the individual differences in training effect, with rs79726572,rs112183859 and rs77187527 explaining 13.7%, 9% and 8%, respectively; the mean GPGS score after weighting was 3.12(range of variation:-3.93 to 27.21). 3) The initial values of prone push, BMI, right upper limb muscle mass, left trunk muscle mass, and GPGS scores were included in the combined model which was constructed by GPGS and phenotype indicators(AUC=0.952, Jorden index=0.767, cut off value 0.251). 4) Bioinformatic analyses showed that the genes or regulated genes underlying 7 SNPs in the genomic prediction model of bench press 1RM training effect were functionally associated with myogenesis; the regulated genes were enriched in Generic Transcription Pathway, Developmental Biology, Myogenesis and other 37 pathways(P<0.01, FDR<0.01). Conclusions: It is the first study to find 7 genetic markers(rs79726572, rs112183859, rs77187527, etc.) related to bench press 1RM training effect based on GWAS. The established “genome-phenotype” comprehensive prediction model has good predictive ability for bench press 1RM training effect. The SNPs related to bench press 1RM training effect are associated with the biological processes of skeletal muscle growth and development, among which the Myogenesis signaling pathway is worth noting.