Abstract A024: AI-Integrated Framework for Predicting ADME and Toxicological Profiles: A Hybrid Data- and Physics-Driven Approach
Bibliographic record
Abstract
Abstract Reliable early-stage prediction of ADME (Absorption, Distribution, Metabolism, and Excretion) and toxicity profiles is a cornerstone of efficient drug discovery. We present an artificial intelligence (AI) assisted, comprehensive, and well-validated platform that integrates machine learning (ML) algorithms with physics-based modelling to predict key pharmacokinetic and safety parameters of small-molecule candidates. The model was developed using deep learning-based Multi-Layer Perceptron (MLP DNN) and Random Forests (RF), trained with the descriptors generated from a proprietary dataset of clinical and non-clinical drug candidates. The MLP DNN captures complex, non-linear patterns, while RF offers interpretability and robustness. The novelty lies in the unique integration of these models with exclusive, real-world drug data, enabling more precise and innovative drug candidate prediction. The robustness of the models was benchmarked with publicly available data as well as open-source ADMET prediction tools. The benchmark study exposed that the tool is comparable to or superior to the existing tools in terms of predictivity. Overall, our tool is capable of demonstrating strong predictive performance across endpoints such as oral bioavailability, blood-brain barrier permeability, CYP450 inhibition, hepatotoxicity, and hERG liability, showing improved consistency across structurally diverse compounds. Case Study: The tool was applied to assess and compare multiple USP1 inhibitors (both known and proprietary) to determine oral absorption, metabolic stability, hepatotoxicity risk and any hERG inhibition potential. Subsequent in vitro studies confirmed these predictions, validating both the accuracy and practical relevance of the model. This work highlights the value of hybrid AI systems that combine data-driven insights with mechanistic understanding. Our approach supports early go/no-go decisions, facilitates lead optimization, and holds promise for streamlining drug development pipelines. Citation Format: Md Ataul Islam, Appaji Baburao. Mandhare, Prashant Bhavar, Uday Surampudi. AI-Integrated Framework for Predicting ADME and Toxicological Profiles: A Hybrid Data- and Physics-Driven Approach [abstract]. In: Proceedings of the AACR Special Conference in Cancer Research: Artificial Intelligence and Machine Learning; 2025 Jul 10-12; Montreal, QC, Canada. Philadelphia (PA): AACR; Clin Cancer Res 2025;31(13_Suppl):Abstract nr A024.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame distilled prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. Learned from the 10,348 direct Codex labels and 10,348 direct Gemma labels. Candidate is the union of thresholded teacher heads; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels or direct frontier model labels.
Codex and Gemma teacher scores by category
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.005 | 0.004 |
| Meta-epidemiology (narrow) | 0.000 | 0.000 |
| Meta-epidemiology (broad) | 0.000 | 0.000 |
| Bibliometrics | 0.000 | 0.000 |
| Science and technology studies | 0.000 | 0.002 |
| Scholarly communication | 0.000 | 0.000 |
| Open science | 0.001 | 0.001 |
| Research integrity | 0.000 | 0.002 |
| Insufficient payload (model declined to judge) | 0.000 | 0.000 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one teacher head, not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".