Stochastic gene expression modeling using the Gillespie algorithm: insights into oral cancer genomic variability

Abstract

Objective: Gene expression is a complex and dynamic process influenced by various factors, particularly in diseases like oral cancer. This study applies the Gillespie algorithm to oral cancer genomic data, emphasizing its utility in exploring gene regulation and variability in tumorigenesis. Material and Methods: The study analyzed gene expression profiles from the NCBI GEO dataset (GSE30784), which includes data from 167 oral squamous cell carcinomas, 17 dysplasia cases, and 45 normal oral tissues. The Gillespie algorithm was employed to simulate stochastic processes governing gene expression, focusing on transcription and degradation reactions. The method involves initializing systems, calculating reaction propensities, and generating time-series data to model the time evolution of gene expression systems. Results: The analysis revealed key insights into transcriptional dynamics, highlighting variability in transcription rates and degradation rates. The study observed a theoretical mean expression level of 5.0 compared to an ensemble mean of 4.7515, indicating stochastic fluctuations. The ensemble Coefficient of Variation (CV) of 0.4125 quantified variability, while the high autocorrelation value (0.8339) indicated that gene expression is significantly influenced by preceding states. These findings provide a normalized measure of gene expression variability and underscore the influence of stochastic processes on cellular systems. Conclusion: The Gillespie algorithm effectively models the stochastic nature of gene expression, uncovering intrinsic noise and variability in oral cancer. By demonstrating the role of transcriptional stochasticity in cellular heterogeneity, this study provides a robust framework for investigating gene regulation in disease contexts, such as cancer progression and drug resistance.

KEYWORDS

Computational biology; Gene expression regulation; Mathematical models; Mouth neoplasms; Stochastic processes.