Artelo Biosciences (Nasdaq: ARTL) has entered a strategic collaboration with AI platform company ScienceMachine to apply machine learning to its fatty acid-binding protein 5 inhibitor program, with ART26.12 — the company's lead small molecule FABP5 inhibitor — as the primary focus. The partnership, which began in 2025, deploys ScienceMachine's AI agent technology against Artelo's internal FABP dataset library to surface novel mechanistic insights, clinical biomarkers, and expansion opportunities across multiple disease areas. Financial terms were not disclosed.
ART26.12 is described by Artelo as the first selective FABP5 inhibitor to enter human clinical studies, currently in Phase I development for chemotherapy-induced peripheral neuropathy. The completed Phase I single ascending dose study, designated ART26.12-100, generated the dataset now being interrogated by ScienceMachine's platform.
Deal context
FABP5 is an intracellular lipid chaperone protein involved in transporting endocannabinoids and other fatty acid-derived signaling molecules within cells. Published preclinical work, including Warren et al. in the Journal of Pain (2024), supports a mechanistic rationale for FABP5 inhibition in neuropathic pain models, particularly oxaliplatin-induced peripheral neuropathy. By blocking FABP5, ART26.12 is designed to modulate downstream lipid-signaling cascades relevant to pain and inflammation without engaging opioid or non-steroidal anti-inflammatory pathways.
ScienceMachine, a privately held San Francisco-based company founded in 2024 and backed by Basis, Opal Ventures, and Vento, operates an AI platform built on autonomous agent architecture. The platform automates data processing, multi-omic analysis, literature synthesis, and report generation, and is applied here to Artelo's proprietary biological datasets rather than to de novo molecular design. In the collaboration, ScienceMachine has identified candidate proteins indicative of FABP5 target engagement in healthy volunteers from the ART26.12-100 study, and has identified novel protein and lipid signatures correlated with ART26.12's dose-responsive analgesic effect in preclinical models.