From Literature to Knowledge Graphs to Drug Repurposing

<p xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance" class="first" dir="auto" id="d392105e255">Network-based drug repurposing requires the seamless synthesis of knowledge from both the scientific literature and curated databases, ensuring that researchers capture the best of human expertise alongside robust computational insights. Yet as the volume of biomedical publications grows exponentially [1], investigators face unprecedented challenges in evidence synthesis and the discovery of novel therapeutic applications for existing drugs [2]. Systematic reviews, guided by frameworks like PRISMA [3], can be thorough but also labor-intensive and prone to human bias [4,5], while advanced computational platforms for disease-module identification frequently remain inaccessible to non-specialists. Consequently, bridging this gap between time-consuming systematic review workflows and powerful, data-driven methods has emerged as a major obstacle in accelerating translational research. <b>We introduce a joint framework, merging Academate—a large language model-powered, end-to-end systematic review assistant —with ChatDRex—a multi-agent, conversation-based platform that harnesses a biomedical knowledge graph and specialized bioinformatics tools. Through retrieval-augmented generation [6], chain-of-thought processing [7], and task-specific agents, this integrated approach eases literature curation, functional enrichment [8], disease-module detection [9], and drug repurposing in an accessible, conversational environment. </b>In testing across multiple domains, Academate demonstrated consistent, moderate-to-substantial agreement with human reviewers during title/abstract and full-text screenings, while ChatDRex enabled robust disease module exploration and rapid drug prioritization workflows. Together, these platforms reduced manual workload, improved reproducibility, and enabled deeper, interactive analyses—particularly in identifying new therapeutic opportunities from vast, scattered datasets.By combining systematic literature review automation with network-based drug repurposing, Academate and ChatDRex furnish a scalable, user-friendly ecosystem that empowers both computational and clinical researchers. Their synergy promotes evidence-based decision-making, accelerates discovery in personalized, and helps translate large-scale biomedical data into targeted, effective interventions. <p xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance" dir="auto" id="d392105e260"> <b>References</b> <p xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance" dir="auto" id="d392105e265">1. Kumar AHS. Rise in polluters of scientific research: How to curtail information pollution (infollution). J Nat Sci Biol Med. 2013;4: 271.2. Pushpakom S, Iorio F, Eyers PA, Escott KJ, Hopper S, Wells A, et al. Drug repurposing: progress, challenges and recommendations. Nat Rev Drug Discov. 2019;18: 41–58.3. Liberati A, Altman DG, Tetzlaff J, Mulrow C, Gøtzsche PC, Ioannidis JPA, et al. The PRISMA statement for reporting systematic reviews and meta-analyses of studies that evaluate healthcare interventions: explanation and elaboration. BMJ. 2009;339: b2700.4. McHugh ML. Interrater reliability: the kappa statistic. Biochem Med (Zagreb). 2012;22: 276–282.5. Hanegraaf P, Wondimu A, Mosselman JJ, de Jong R, Abogunrin S, Queiros L, et al. Inter-reviewer reliability of human literature reviewing and implications for the introduction of machine-assisted systematic reviews: a mixed-methods review. BMJ Open. 2024;14: e076912.6. Lewis P, Perez E, Piktus A, Petroni F, Karpukhin V, Goyal N, et al. Retrieval-augmented generation for knowledge-intensive NLP tasks. arXiv [cs.CL]. 2020. Available: http://arxiv.org/abs/2005.114017. An intelligent LLM-powered personalized assistant for digital banking using LangGraph and chain of thoughts.8. Adamowicz K, Maier A, Baumbach J, Blumenthal DB. Online in silico validation of disease and gene sets, clusterings or subnetworks with DIGEST. Brief Bioinform. 2022;23. doi:10.1093/bib/bbac2479. Ghiassian SD, Menche J, Barabási A-L. A DIseAse MOdule Detection (DIAMOnD) algorithm derived from a systematic analysis of connectivity patterns of disease proteins in the human interactome. PLoS Comput Biol. 2015;11: e1004120.