AI- and LLM-Driven Drug Discovery Pipeline for Predictive Evaluation of Nimbolide and Salannin in Oral Squamous Cell Carcinoma

Abstract

Background: Oral squamous cell carcinoma (OSCC) is a major malignancy associated with complex molecular pathways involving inflammation, oxidative stress, immune dysregulation, and abnormal cell proliferation. Conventional drug discovery approaches are often time-consuming and less effective for multi-target diseases. Recent advances in artificial intelligence (AI) and large language models (LLMs) have enabled systematic exploration of phytochemicals such as neem-derived compounds.
Objective: To develop an AI- and LLM-driven integrative research pipeline using the Swalife platform to evaluate the therapeutic potential of neem bioactives, particularly nimbolide and salannin, in OSCC.
Methods: A multi-stage AI-assisted workflow was implemented using literature mining from PubMed, Google Scholar, and ScienceDirect, followed by LLM-assisted data extraction using Perplexity AI. Six modules were applied: target identification, lead optimization, in vitro design, in vivo design, clinical pharmacovigilance, and market/IPR analysis. Outputs were structured into HTML datasets and integrated into predictive and preventive medicine frameworks.
Results: Nimbolide and salannin showed multi-target activity across major OSCC pathways, including NF-κB, STAT3, PI3K/Akt, and EGFR. Approximately 70–80% target overlap was observed between neem bioactives and OSCC molecular pathways. Predictive modeling confirmed strong anti-inflammatory, pro-apoptotic, anti- angiogenic, and immunomodulatory effects.
Conclusion: The AI- and LLM-driven Swalife pipeline effectively bridges mechanistic discovery and predictive medicine. Nimbolide and salannin demonstrate strong potential as multi-target adjunct therapies in OSCC, supporting further experimental and clinical validation.