Fig. 5
From: Are we there yet? A machine learning architecture to predict organotropic metastases
Shared significantly overrepresented biological processes. Gene set enrichment analysis was conducted using the clusterProfiler package in R. The Go ontology database was used to investigate feature enrichment in Biological Processes for each metastatic location in each cancer type that was classified by the model. SimplifyEnrichment package was used to cluster the semantic similarity between shared overrepresented biological processes in tumors metastasizing to concordant locations. A Enriched processes in Bone metastases. B Enriched processes in Liver metastases. C Enriched processes in Lung metastases. D Enriched processes in Lymph Node metastases. Statistical significance and GO:ID enrichment results included in Additional file 1: data tables. Similarity scores are on a scale of 0 to 1