DCPM-ADMET: Fusion of Dual-channel Pre-trained Model and Molecular Fingerprints to enhance Drug ADMET Properties Prediction

Yuchen Zeng, Yue Qi, Leilei Zhang, Kaili Jiang, Xiaofei Zhou, Liang Lü, Jianping LinResearch Square 2025 · DOI

Abstract

The absorption, distribution, metabolism, excretion, and toxicity (ADMET) properties of drugs are critical to their efficacy and safety in clinical trials; however, traditional machine learning methods have limited generalization ability in ADMET prediction due to insufficient data. To address this issue, we developed DCPM-ADMET, an innovative pre-trained model with higher accuracy, whose architecture employs a two-channel system—including an XLNet-based module for capturing the semantics of molecular sequences, an RNN-based component for small molecule property extraction, and ECFP fingerprints for capturing molecular substructures—and after initial pre-training, the model outperforms traditional methods in prediction accuracy on multiple benchmark datasets for molecular properties; additionally, we fine-tuned it on a self-constructed database containing 465,470 entries covering 97 ADMET properties, and by integrating these 97 prediction models and 36 computational properties, we further developed a free online ADMET prediction tool with 133 endpoints (available at http://admet.bioai-global.com/), which is designed to assist researchers in conducting comprehensive molecular ADMET predictions.

Introduction

In the high-stakes field of drug discovery, potency is only half the battle. Approximately 40–60% of drug candidates fail during clinical trials due to poor ADMET properties—instances where the body cannot safely process the drug or the drug fails to reach its target effectively. Predicting these properties early using machine learning is a primary objective for the industry, yet it faces a significant "data desert" problem. Traditional models often struggle to generalize because experimental ADMET data is scarce and noisy. Furthermore, many current AI models suffer from a "pre-training discrepancy," where the learning phase (e.g., masking tokens) does not align with real-world application (where data is complete).

Methodology: The Dual-Channel Architecture

The researchers introduced DCPM-ADMET (Dual-channel Pre-trained Model), which utilizes a sophisticated three-pronged approach to molecular understanding:

  1. Channel 1 (XLNet-based): Unlike standard BERT-like architectures that rely on masking, this channel uses XLNet’s permutation-based autoregressive strategy. This allows the model to learn bidirectional context without artificial masks, effectively solving the pre-training discrepancy.
  2. Channel 2 (RNN-based): This channel focuses on chemical logic through a Sequence-to-Sequence (Seq2Seq) translation from SMILES to InChI identifiers, while simultaneously regressing nine fundamental physicochemical descriptors.
  3. Feature Fusion: The model integrates these deep learning features with ECFP (Extended Connectivity Fingerprints). By explicitly encoding molecular substructures, the model ensures it retains critical information about chemical fragments that dictate toxicity and reactivity.

Data Scale and Benchmarking

The scale of data utilized in this study represents a significant advancement over previous efforts. The model underwent initial pre-training on a massive set of 111,295,885 data points from PubChem. For specific ADMET tasks, the researchers constructed a proprietary database of 465,470 entries covering 97 properties, aggregated from high-quality sources like ChEMBL and the Therapeutics Data Commons (TDC).

Comparison of DCPM-ADMET with other public ADMET prediction platforms, highlighting its superior data size and endpoint coverage.

Key Results and Performance

DCPM-ADMET demonstrated superior predictive accuracy across multiple benchmark datasets. In the MoleculeNet benchmark, the model outperformed traditional ECFP-based methods in 7 out of 10 tasks. Specifically, in regression tasks such as Caco-2 permeability, the model achieved a Pearson Correlation Coefficient (PCC) of 0.82, indicating a high degree of alignment between predicted and experimental values.

Predictive performance (ROC-AUC) of DCPM versus ECFP on classification tasks from the MoleculeNet benchmark.

Beyond benchmark metrics, the model showed real-world utility by correctly identifying the toxicity of 80% of a test set containing 45 drugs withdrawn from the market. For example, it flagged the drug Nefazodone as high-risk for liver injury, catching subtle toxicological signals that traditional screenings often miss.

Conclusion and Future Directions

The development of DCPM-ADMET addresses the fundamental limitation of poor generalization in computational pharmacology. By deploying their findings into a free online tool offering 133 endpoints—the highest number among comparable public platforms—the researchers have provided an essential resource for early-stage drug discovery. Future work will focus on extending the DCPM architecture to predict drug-drug interactions (DDIs) and moving from property prediction to the de novo design of molecules with optimized ADMET profiles.

Summary updated

Abstract

<title>Abstract</title> The absorption, distribution, metabolism, excretion, and toxicity (ADMET) properties of drugs are critical to their efficacy and safety in clinical trials; however, traditional machine learning methods have limited generalization ability in ADMET prediction due to insufficient data. To address this issue, we developed DCPM-ADMET, an innovative pre-trained model with higher accuracy, whose architecture employs a two-channel system—including an XLNet-based module for capturing the semantics of molecular sequences, an RNN-based component for small molecule property extraction, and ECFP fingerprints for capturing molecular substructures—and after initial pre-training, the model outperforms traditional methods in prediction accuracy on multiple benchmark datasets for molecular properties; additionally, we fine-tuned it on a self-constructed database containing 465,470 entries covering 97 ADMET properties, and by integrating these 97 prediction models and 36 computational properties, we further developed a free online ADMET prediction tool with 133 endpoints (available at http://admet.bioai-global.com/), which is designed to assist researchers in conducting comprehensive molecular ADMET predictions. Scientific contribution The development of DCPM-ADMET represents a seminal advancement in computational pharmacology. This novel pre-trained model successfully addresses the fundamental limitation of poor generalization in predicting ADMET (Absorption, Distribution, Metabolism, Excretion, and Toxicity) properties, a challenge stemming from data insufficiency in traditional machine learning approaches. Our architecture innovatively employs a dual-channel system: an XLNet-based module for deep capture of molecular sequence semantics, an RNN-based component for efficient extraction of small molecule properties, and ECFP fingerprints to comprehensively encode structural features. Following intensive pre-training, DCPM-ADMET demonstrates superior predictive accuracy across multiple benchmark molecular property datasets. Furthermore, we fine-tuned this model on a proprietary, large-scale database of 465,470 entries covering 97 ADMET endpoints. By integrating the resultant 97 prediction models with 36 calculated physicochemical properties, we have deployed a free, high-throughput online ADMET prediction tool with 133 endpoints, which is set to become an essential resource for guiding early-stage drug discovery and safety assessment.
BibTeX
@inproceedings{zeng2025dcpm,
  author = {Yuchen Zeng and Yue Qi and Leilei Zhang and Kaili Jiang and Xiaofei Zhou and Liang Lü and Jianping Lin},
  title = {{DCPM-ADMET: Fusion of Dual-channel Pre-trained Model and Molecular Fingerprints to enhance Drug ADMET Properties Prediction}},
  booktitle = {Research Square},
  year = {2025},
  doi = {10.21203/rs.3.rs-8326331/v1},
}
Figure