Detection of driver mutations and genomic signatures in endometrial cancers using artificial intelligence algorithms

Anda Stan ¹, Korey Bosart ¹, Mehak Kaur ¹

⁰Biology Program, The Ohio State University, Marion, Ohio, United States of America.

Plos One +

|

February 26, 2024

View abstract on PubMed

Summary

This study analyzed endometrial cancer genomes, identifying key driver mutations using artificial intelligence. These drivers, comprising about 2.5% of mutations, are crucial for cancer development and offer new diagnostic and therapeutic targets.

Area Of Science

Genomics
Oncology
Bioinformatics

Background

Endometrial cancer (EC) genomic analysis reveals molecular signatures for classification and prognostication.
Artificial intelligence (AI) aids in distinguishing driver mutations from passenger mutations based on various parameters.

Purpose Of The Study

To classify all mutations in endometrial cancer genomes as driver or passenger.
To characterize nucleotide-level mutation signatures, chromosomal rearrangements, and gene expression profiles in EC.
To investigate the impact of driver mutations on protein structure using in silico analysis.

Main Methods

Analysis of endometrial cancer genomes from the Catalogue of Somatic Mutations in Cancers (COSMIC).
Application of artificial intelligence algorithms for mutation classification.
Characterization of nucleotide substitution signatures, chromosomal rearrangements, and gene expression.
In silico protein structure analysis of driver mutations in CTNNB1 and TP53.

Main Results

Approximately 2.5% of all mutations were identified as drivers, contributing to cellular transformation and immortalization.
Endometrial cancers exhibit distinct nucleotide substitution and chromosomal rearrangement signatures compared to other cancers.
High expression of CLDN18 was observed, implicating it in cancer growth and metastasis.
Specific mutations in CTNNB1 and TP53 were found to increase protein stability, potentially driving cellular transformation.

Conclusions

AI algorithms effectively predict key cancer drivers.
This study identified novel molecular signatures in endometrial cancer.
Findings enhance understanding of EC pathogenesis and improve tools for classification, diagnosis, and treatment.