About The Tool
We present a novel web-based bioinformatics tool designed to facilitate the identification of novel therapeutic targets and biomarkers for drug discovery. The tool integrates multi-omics datasets rooted in human genetics and utilizes experimentally validated protein-protein interaction (PPI) networks to perform genome-wide analyses of proteins that physically interact with genes responsible for disease phenotypes. A key feature of this tool is its real-time large-scale data processing capability, enabled by its efficient architecture and cloud-based framework. Additionally, the tool incorporates an integrated large language model (LLM) to assist scientists in exploring biological insights from the generated network and multi-omics data. The LLM enhances the interpretation and exploration of complex biological relationships, offering a more interactive and intuitive analysis experience. This integration of multi-omics data, PPI networks, and AI-driven exploration provides a powerful framework for accelerating the discovery of novel drug targets
Tool Architecture
The tool consists of three core components: a knowledge graph, which maps relationships between pathways, targets, and diseases; a knowledge base, which collects multi-omics data information from various sources; and a knowledge bot, which responds with the exact compact answer useful for the analytics along with relevant references and citations. Below is the high-level architecture diagram, explaining the same:
