Fig. 2

Unsupervised Clustering Analysis Across Multiple Cohorts Reveals a Strong Association Between Abundant Immune Cell Infiltration Regulated by Key Gene Modules and Adverse Prognosis in IPF Patients. (A) Unsupervised clustering and t-SNE dimensionality reduction analyses were performed on IPF patients from the GSE70866 dataset based on immune cell infiltration scores. (B) Heatmaps displaying immune cell infiltration scores for Subtypes A and B in the GSE70866 dataset. (C) Unsupervised clustering and t-SNE dimensionality reduction analyses were conducted on IPF patients from the GSE10667 dataset based on immune cell infiltration scores. (D) Heatmaps displaying immune cell infiltration scores for Subtypes A and B in the GSE10667 dataset. (E) Unsupervised clustering and t-SNE dimensionality reduction analyses were performed on IPF patients from the GSE110147 dataset based on immune cell infiltration scores. (F) Heatmaps displaying immune cell infiltration scores for Subtypes A and B in the GSE110147 dataset. (G) Survival analysis of patients in Subtypes A and B in the GSE70866 dataset. (H) Soft threshold determination. (I) Identification of gene clustering modules. (J) Correlation analysis between gene modules and phenotypes, with pink modules exhibiting a high correlation with subtypes. (K) Correlation coefficient analysis between gene significance (GS) and module membership (MM). (L) KEGG enrichment analysis. (M) GO enrichment analysis. (N) KEGG enrichment analysis.