A hybrid framework integrating structural machine learning and 3D liver-on-chip assay for drug-induced liver injury prediction

Fan Zhang, Yu Zhou, Duanchen Ding, Feng Zhang, Rong-Rong Xiao, Xiaoni Ai

Abstract Drug-induced liver injury (DILI) remains a major cause of clinical attrition and postmarketing withdrawal, but structure only DILI predictors are difficult to compare because public benchmarks are vulnerable to compound overlap, scaffold similarity and shared label provenance. We present Oakuloid TM , an open DILI prediction framework that pairs a leakage audited structure based model with an optional iBAC 3D primary human hepatocyte IC 50 /C max confirmation signal. The structural model integrates gradient boosted descriptor backbones, fingerprint random forests and LivTox proxy-DILI features through a logistic meta learner. Its evaluation is designed as part of the contribution: internal DILIrank, strict external TDC, scaffold disjoint TDC and independent Geci provenance checks are reported with released per compound predictions. Oakuloid reaches AUROC 0.811 on the strict external TDC benchmark and remains competitive under scaffold and fully clean TDC filtering. A channel attribution ablation shows that the external benchmark lead is driven by descriptor based gradient boosted trees rather than by DILIPredictor derived proxy features, reducing a potential circularity concern. The wet lab IC 50 /C max signal is largely orthogonal to structure and supports a confirmation mode that shifts the internal operating point toward higher specificity without claiming a universal AUROC gain. Oakuloid is released with code, model artifacts, calibration analysis, a 122 compound wet lab benchmark and a model card under the Apache License 2.0, supporting reproducible DILI screening and benchmark auditing.