Molecular Network-Based Drug Prediction in Thyroid Cancer

Area of Science:

• Oncology
• Genomics
• Pharmacology

Background:

• Thyroid cancer lacks effective treatments, necessitating improved drug selection strategies.
• Existing methods like the Connectivity Map (CMAP) have limitations, including reliance on cell lines and ignoring gene regulation.
• A comprehensive analysis of thyroid cancer perturbations at gene expression, co-expression, and module levels is needed.

Purpose of the Study:

• To develop and apply a novel drug selection pipeline for thyroid cancer.
• To identify drugs capable of reversing gene expression and co-expression changes in thyroid cancer tissues.
• To explore the relationship between thyroid hormone secretion and thyroid cancer proliferation.

Main Methods:

• Comprehensive analysis of thyroid cancer perturbations using The Cancer Genome Atlas (TCGA) data (56 normal, 500 cancer samples).
• Assessment of gene expression, gene co-expression, and gene module variations.
• Development of a drug selection pipeline based on tissue-derived drug signatures and Fisher's exact test.

Main Results:

• Identified 812 up-regulated and 213 down-regulated genes, enriched in extracellular matrix and synaptic functions.
• Discovered 33,778 differentiated co-expressed gene pairs forming a module linked to impaired immunity.
• Predicted drugs including baclofen, nevirapine, and glucocorticoids that could reverse cancer-induced changes.

Conclusions:

• The developed pipeline effectively identifies potential drugs for thyroid cancer by analyzing tissue-level perturbations.
• Gene co-expression analysis reveals immune dysfunction associated with thyroid cancer.
• Thyroid hormone secretion regulation may be a key factor in inhibiting thyroid cancer cell proliferation.