Development of a data colletion and analysis tool for protein - ligand interactions

Abstract

Analysis of protein – ligand interactions guides the development of new drugs. For protein - ligand interaction studies, first step is the construction of an accurate dataset. This data collection process can be completed either by manual search in databases or by using computer-assisted data collection methods. Manual data collection is difficult, time consuming and prone to errors. In this work, we present a novel tool to collect protein-ligand interaction data. We first introduce a protein – ligand interaction data collection tool using UniProt, ChEMBL, PubChem, PDB and BindingDB as its source databases. In the second part, we use this tool to analyze protein – ligand interactions of sphingolipid and insulin metabolisms. First, the datasets of both metabolisms were constructed, then their ligand centric network models were built for ligand analysis. Based on these networks, first the interactions within sphingolipid metabolism proteins, then their interactions with insulin proteins were analyzed. According to the ligand analysis, specific interactions and significant drugs were highlighted. Besides promiscuous drugs interacting with too many proteins, Tamoxifen and Altretamine cancer drugs interacted with key sphingolipid proteins, namely GLCM, ARSA and AGAL. Ceritinib, used for the treatment of nonsmall cell lung cancer, interacted with Kit and Lyn kinases. This ligand based interaction network analysis highlighted the synergy between these two networks.