Additional file 1 of Spatial immunogenomic patterns associated with lymph node metastasis in lung adenocarcinoma

Additional file 1: Table 1. Preoperative clinicopathologic and genomic features associated with pathologic LN metastasis in PKPH NGS cohort. Table 2.1. Patterns of co-occurrence and mutual exclusivity in the PKPH NGS cohort. Table 2.2. Co-occurrence and mutual exclusivity patterns in the pN positive patients of the PKPH NGS cohort. Table 2.3. Co-occurrence and mutual exclusivity patterns in the pN negative patients of the PKPH NGS cohort. Table 3. Glossary. Table 4. Clinicopathologic characteristics of the PKTOI cohort. Figure 1. Univariable and Multivariable Logistic Regression Analysis. A: Univariable logistic regression analysis was performed on preoperative clinicopathologic and genomic features associated with pathologic LN metastasis in the NGS cohort. B: Multivariable logistic regression analysis was conducted on preoperative clinicopathologic and genomic features associated with pathologic LN metastasis in the NGS cohort. Variables with p < 0.05 were highlighted with notable markers. Figure 2. COME Analysis of Oncogenic Pathways. A-C: Co-occurrence (red) and mutual exclusivity (blue) patterns of driver genes were analyzed in the entire cohort, as well as in pN positive and negative groups. D-F: COME analysis of the mitotic pathway was conducted in our whole cohort, pN negative, and pN positive groups. G-I: COME analysis of the mitotic pathway was performed in the MSK cohort, focusing on pN negative and pN positive groups. Abbreviation: COME, co-occurrence and mutual exclusivity. Figure 3. Cellular Metacluster Densities According to the Cellular Immunologic Distribution in LUAD Primary Tumors (n = 92). A, B: The distribution of the numbers of cellular connections and pairs of cell distances among the cellular metaclusters was analyzed. C, D, E: The cell distances (µm) between diverse cellular metacluster pairs were compared: * p < 0.05, ** p < 0.001. Data were presented as means ± SEMs. Statistical analysis was conducted using the Mann-Whitney U test. Figure 4. Genetic Features Analysis of TIME Patterns. A: The silhouette coefficient of the TIME subtypes was calculated and presented. B: Co-occurrence gene analysis was performed for the TIME subtypes. C: The mutation frequencies of genes were compared among the TIME subtypes. Abbreviation: TIME, tumor immune microenvironment. Figure 5. Identification of the Immunogenomic Features According to CN Distribution in the Primary Tumor. A: The distribution of CN profiles was analyzed across the PKPH mIHC cohort (n = 92). B: Statistical analysis showed a positive correlation between epithelial CNs and the proportion of epithelial metaclusters (p < 0.05). No significant difference was observed in the epithelial-enriched CNs (Epithelial-CN 2, 5, 6) between pN negative and positive primary tumor groups (p > 0.05). C: Statistical analysis revealed a positive correlation between CNs enriched in CD8+ and CD4+ T cells (CN 1) and the proportion of T cell metaclusters (p < 0.05). No significant difference was observed in CN 1 between pN negative and positive primary tumor groups (p > 0.05). D: Immunogenomic patterns were integrated with genetic alterations and CNs in our PKPH cohort. Abbreviations: CN, cellular neighborhoods; PKPH, Peking University People’s Hospital. Figure 6. ImGene Model Predicted the LN Stage. A, B, C: The assessment indices of the ImGene multimodel, using genetic and mIHC data, exhibited superior performance compared to the ImFeatures monomodels on both the training and validation sets. D: The ROC curve of the PKU and FUDAN N2 stage prediction model was presented. Abbreviations: PKPH, Peking University People’s Hospital; PKTOI, Peking University People’s Hospital Thoracic Oncology Institution; ROC, receiver operating characteristic; PKU, Peking University.