Table 1.
Baseline Characteristics of the participants.
Table 2.
Physiological adaptations to 12 weeks of training.
Fig 1.
Boxplots of raw counts and VST-transformed expression data.
Panels A and B display the boxplots of raw counts (log₁₀-scaled) and VST-transformed expression data, respectively. The boxplots depict the data distribution across samples, represented by the interquartile range (box), median (central line), whiskers, and outlier points.
Fig 2.
Volcano plots and heatmaps of DEGs.
(A-C) Upper panels: Volcano plots displaying the distribution of DEGs. The y-axis represents -log₁₀(adjusted p-value), and the x-axis represents log₂(fold change). Red dots denote significantly upregulated genes, blue dots denote downregulated genes, and gray dots represent non-significant genes. The top 10 most significant genes (by adjusted p-value) are labeled in each plot. (D-F) Lower panels: Heatmaps illustrating the relative expression levels (Z-score) of DEGs across Pre- and Post-intervention samples. The color gradient from blue to red indicates low to high expression, highlighting distinct expression patterns among the different training groups.
Fig 3.
Functional enrichment analysis of DEGs.
(A-C) Results for the CET, RES, and END groups, respectively. For each panel, the left side presents a Sankey diagram, illustrating the linkages between genes and their significantly enriched pathways. The right side displays a dot plot, where the y-axis represents the enrichment terms, and the x-axis represents the GeneRatio. Dot color corresponds to the statistical significance (-log₁₀(Pvalue)), with a gradient from blue (less significant) to red (more significant). Dot size is proportional to the number of genes associated with each term.
Fig 4.
(A-B) CET group, (C-D) RES group, (E-F) END group. Each group displays up to five significantly activated and five inhibited pathways (fewer are shown if the threshold is not met). Bubble size corresponds to the number of enriched genes (Count) in the gene set. The color gradient represents the adjusted p-value (p.adjust), with red indicating higher significance and blue indicating lower significance. The horizontal axis shows the Normalized Enrichment Score (NES), where positive values denote pathway activation and negative values denote inhibition.
Fig 5.
Identification of differential modules by WGCNA pre- and post-exercise across groups.
(A–B) Soft threshold screening results: Based on the scale-free topology fit index and mean connectivity, a soft threshold power β = 9 was selected as the optimal value, achieving an approximate scale-free network topology (R² ≈ 0.9). (C) Gene co-expression clustering dendrogram and module assignment: Different colored bars below the dendrogram represent the 20 identified gene modules. (D) Heatmap displaying the difference in module eigengene (ME) values (Post − Pre) and their significance for each module across the CET, RES, and END groups. Color intensity represents the effect size (eigengene value change), with corresponding p-values indicated in parentheses.
Fig 6.
Screening of candidate key genes and PPI network construction.
(A) Venn diagram illustrating the overlap between genes from the MEbrown module and the DEGs from the CET and END groups. (B–D) Topological layout of the PPI networks for candidate key genes in the (B) CET, (C) RES, and (D) END groups, respectively. Node color intensity corresponds to the Degree centrality, with red indicating higher degree and yellow indicating lower degree.
Fig 7.
Expression levels and correlations of core genes across groups.
(A–C) Comparison of VST-normalized relative expression levels for the key genes before (Pre) and after (Post) exercise in the (A) CET, (B) RES, and (C) END groups, respectively. Each subpanel combines a violin plot with an overlaid boxplot, and individual lines connect paired measurements from the same subject. Statistical significance was assessed using the paired Wilcoxon test. (D–F) Spearman correlation heatmaps illustrating pairwise correlations among these core genes within the (D) CET, (E) RES, and (F) END groups. The correlation coefficient (r) is displayed above each cell, with the corresponding p-value indicated in parentheses below. The background color intensity reflects the strength of the correlation. *P ＜ 0.05, **P ＜ 0.01.
Fig 8.
TF–mRNA–miRNA regulatory networks.
(A-C) Networks for the CET, RES, and END groups, respectively. Red nodes represent mRNAs, green nodes represent predicted transcription factors (TFs), and purple nodes represent regulatory miRNAs. Edges display predicted regulatory relationships between TFs and mRNAs, as well as between miRNAs and mRNAs.
Fig 9.
(A) Stacked bar plot showing the relative proportions of muscle fiber types (slow-twitch and fast-twitch fibers) for all samples. Each bar represents one sample, with the total height normalized to 100%. The colored segments within each bar indicate the percentage of slow-twitch (blue) and fast-twitch (red) fibers. (B–C) Violin plots displaying the paired changes in (B) slow-twitch and (C) fast-twitch fiber proportions before and after intervention in the three groups (CET, RES, END). The outlines show the distribution of fiber percentage values within each group, while lines connect paired measurements (pre- vs. post-intervention) from the same subject. Statistical significance was assessed using the paired Wilcoxon test.