Pharmacological assessment of compounds as potential therapeutics for cervical cancer.

MOTIVATION

Cervical cancer remains a leading cause of gynecological mortality, with existing treatments often limited by resistance and suboptimal efficacy. While is rich in phytochemicals with reported anticancer properties, their relevance to cervical cancer-specific molecular targets remains underexplored. Here, we integrated transcriptomic profiling, cheminformatics, and survival modeling to evaluate the therapeutic potential of compounds in cervical cancer.

RESULTS

From 158 bioactive compounds with favorable pharmacokinetic and drug-likeness properties, we predicted gene targets and intersected them with 1779 differentially expressed genes identified from bulk RNA-sequencing of 304 cervical cancer tumors and 47 normal cervical tissues. This yielded 43 gene targets that were significantly dysregulated in cervical cancer. Pathway enrichment revealed involvement in tumorigenesis, immune modulation, and cell cycle regulation, with fold enrichment computed as the ratio of observed-to-expected gene overlap. Survival analysis identified 14 of these genes as markers of poor prognosis, with matrix metalloproteinase-7 (MMP7) emerging as an independent prognostic marker of adverse outcome. A Random-Forest-Regression model trained on 499 experimentally validated MMP7 inhibitors identified carnosol-a compound-as a top-ranking candidate with high predicted activity. These findings nominate carnosol as a promising therapeutic lead for cervical cancer and lay the groundwork for future experimental validation.

AVAILABILITY AND IMPLEMENTATION

The data supporting the findings of this study, including bulk RNA-seq gene expression data, survival, and phenotype data, are available through the TCGA database. These data can be accessed via the Xenabrowser platform (https://xenabrowser.net) using the reference identifier [TCGA Cervical Cancer (CESC)]. Corresponding healthy cervical tissue RNA-seq data, are available through the Genotype-Tissue Expression (GTEx) project (https://www.gtexportal.org/home/). The codes used for differential gene expression (DGE) analysis, pathway enrichment, and survival analysis, as well as scripts for generating volcano plots (DGE analysis), Kaplan-Meier survival plots, and boxplots (gene expression), and machine learning implementations are available on GitHub (https://github.com/Ponaskillzyy/Coffea_arabica_Potential_in_Cervical_Cancer).