pyTARG

pyTARG predicts potential metabolic drug targets and quantifies the effects of metabolic inhibition on cancer cell proliferation by integrating Genome Scale Metabolic Models (GSMMs) with RNA-seq data and chemical similarity searches between KEGG metabolites and DrugBank compounds.

Key Features:

• Integration with GSMMs: Uses human Genome Scale Metabolic Models constrained with RNA-seq data to reflect cell-specific metabolic flux distributions in cancerous and healthy cells.
• Drug Target Identification via chemical similarity: Matches KEGG-indexed human metabolites to DrugBank compounds and computes Tanimoto scores to predict compounds likely to bind metabolic enzymes.
• Quantitative analysis of metabolic fluxes: Quantifies the effects of inhibiting specific metabolic reactions on cellular metabolism and proliferation, enabling comparison of differential drug responses (e.g., MCF7 breast cancer cells versus ASM airway smooth muscle cells).
• Identification of therapeutic windows and pathway prioritization: Analyzes RNA-seq data from 34 cancer cell lines and 26 healthy tissues to identify targetable metabolic pathways, highlighting the mevalonate pathway and its role in cholesterol synthesis in cells lacking transporters such as NPC1L1 and LPL.

Scientific Applications:

• Cancer drug discovery: Predicts candidate metabolic inhibitors and prioritizes compounds for targeting cancer-specific metabolic vulnerabilities.
• Therapeutic pathway prioritization: Identifies and ranks metabolic pathways, such as the mevalonate pathway, as potential anticancer targets based on modelled metabolic impacts.
• Precision oncology: Supports selection of metabolic targets tailored to specific cancer and healthy tissue RNA-seq profiles to maximize selectivity.

Methodology:

Constraining GSMMs with RNA-seq data; computing Tanimoto structural similarity between KEGG metabolites and DrugBank compounds to predict drug–metabolite interactions; simulating the impact of metabolic reaction inhibition on cell proliferation and metabolic fluxes using GSMM simulations.