Identification of Protein Markers for Chronic Ischemic Heart Disease Through Integrated Analysis of the Human Plasma Proteome and Genome-Wide Association Data

Figures

• 

  Figure 1. Flowchart of the research design. The core objective is to eliminate reverse causal interference through bidirectional MR analysis, while integrating pathway enrichment, database validation, and drugability assessment to ensure that the selected protein markers possess both “causal association” and “clinical translational value”. Some steps (such as instrumental variable selection and MR analysis) have been repeatedly verified to reduce bias.

• 

  Figure 2. Manhattan plot for association screening based on the directional consistency results of MR analysis and five methods (the red horizontal line represents the significance threshold of p = 0.05, and the genes highlighted in red indicate significant associations). The x-axis (Chr) represents the chromosome numbers of the 22 human autosomes (each in a distinct color), indicating the genomic positions of the protein-coding genes(). The y-axis (−log10 p) indicates the statistical significance of the association between each protein and CIHD, with higher values reflecting stronger evidence of association. (Note: This figure is based on results from three independent MR analyses. All significantly associated proteins highlighted in red (p < 0.05) fulfilled the criteria of consistent effect direction across all five methods and no significant pleiotropy in each of the three repetitions. The observed results demonstrate stability and reliability).

• 

  Figure 3. Reverse MR scatter plot: This figure presents the reverse MR analysis for four proteins (ROR1, S100A6, SHANK3, CFHR2), with CIHD as the exposure and each protein as the outcome. The analysis includes four subplots, each corresponding to one of the proteins in sequence. The x-axis represents the genetic effect of SNPs on CIHD (β value); the y-axis represents the genetic effect of SNPs on the respective protein (β value). The lines are colored to distinguish results from five MR methods (e.g., blue: IVW method; red: weighted median method).

• 

  Figure 4. Forest plot of the association between exposure and outcome using the IVW method for protein markers (showing the number of SNPs, p values, OR value, and 95% confidence intervals for each exposure factor). Horizontal axis (OR): Ratio of causal effect (OR < 1 = protective factor, OR > 1 = risk factor); Black dotted line: Zero effect line (OR = 1) (the blue origin represents a protective factor, the red dot represents a risk factor). (Note: The OR values and 95% confidence intervals in the figure represent the mean estimates from three independent IVW analyses; the number of SNPs for each exposure factor corresponds to the intersection of twice pQTL screenings).

• 

  Figure 5. Barplot of KEGG enrichment analysis results. Y-axis: Name of the KEGG pathway; X-axis: Number of candidate proteins contained in the pathway; The length of the column chart is based on the quantity, while the color depth is based on the p value. (Note: All the pathways marked in this figure were obtained through three independent repeated KEGG enrichment analyses.).

• 

  Figure 6. Heatmap of NGDC retrieval results (displays the associations between proteins and diseases/traits, as well as the distribution of average colocalization scores; gray represents score = 0). Vertical axis (Protein): Lists the names of all 23 associated protein markers (such as ADH5, CXCL12, CD14, etc.); Horizontal axis (Score Type): Includes two types of scores—① “Disease score (disease score)”: Reflects the known association strength of the protein with CIHD and related traits (such as blood lipids, blood pressure), calculated based on the cardiovascular disease map (CVD-A) in the NGDC database for the European population; ② “Coloc score (co-localization score)”: Reflects the genetic co-localization probability of the gene locus corresponding to the protein and the susceptibility loci of CIHD.

• 

  Figure 7. Cnetplot of protein-drug enrichment relationships (showing the relationships between 11 protein markers and 10 enriched drugs). Node type: Blue nodes: Represent candidate protein markers; Orange nodes: Represent enriched drugs/compounds selected from the DSIGDB database; Connection meaning: The black lines connecting proteins and drugs indicate a significant enrichment association between the drug and the protein (hypergeometric test p < 0.1, corrected p < 0.1); Key example: PLAU (urokinase-type plasminogen activator) is associated with 6 drugs (the most extensive association), while CXCL12 is associated with 4 drugs, suggesting that these proteins have a high potential for drug targeting.

• 

  Figure 8. Charts of the retrieval results from databases for protein markers/targets (showing the number of related compounds of 28 protein markers, the approved drugs or the drugs currently in clinical trials, as well as the corresponding clinical trials). Horizontal axis (Target): Lists all 28 candidate protein biomarker names; Vertical axis (Value): Corresponds to the quantity of three types of data—“ChEMBL_Associated Compounds”: The number of small molecule compounds that bind to proteins in the ChEMBL database (version 35); “Approved Drugs and Clinical Candidates”: The number of approved drugs or clinical-stage candidate drugs; “Clinicaltrials”: The number of clinical trials on the ClinicalTrials website that target or use this protein as a biomarker.