Personalized oncology in pheochromocytomas and paragangliomas: integrating genetic analysis with machine learning.

Abstract

Pheochromocytomas and paragangliomas (PCCs/PGLs) are uncommon neuroendocrine tumors with a significant genetic tendency. Approximately 35-40% of these tumors are associated with genetic factors. The present study performed a thorough analysis using publicly accessible genetic and clinical data from the Cancer Genome Atlas (TCGA) to examine the involvement of six genes, namely GBP1, KIF13B, GPT, CSDE1, CEP164, and CLCA1, in the development of PCCs/PGLs. By employing multi-omics data, this study investigates the relationship between mutational patterns and the prognosis of tumors, focusing on the possibility of tailoring treatment methods to individual patients. The study utilizes Mutect2 to detect somatic mutations with high confidence in whole-exome sequencing data from PCCG samples. The study uncovers mild effects on protein function caused by particular mutations, including GBP1 (p.Cys12Tyr), KIF13B (p.Arg847Gly), and GPT (p.Gln50Arg). A random forest classifier uses mutational profiles to predict potential drug recommendations, proposing a focused therapy strategy. This study thoroughly analyzes the genetic mutations found in PCCs/PGLs, highlighting the significance of precision medicine in developing specific treatments for these uncommon types of cancer. This study aims to improve the understanding of the development of tumors and identify personalized treatment approaches by combining genetic data with machine learning analyses.