Loading
Life sciences · Preprint
arXiv · September 9, 2026
Posted before peer review. The findings may change or fail to hold.
This preprint describes development of an XGBoost-based machine learning classifier trained on 2,039 molecular structures to predict blood-brain barrier permeability from 15 physicochemical descriptors. The model achieved 88.97% accuracy and 0.9282 ROC-AUC on internal cross-validation, identifying TPSA, HBD, and LogP as the most predictive features. However, as an unreviewed computational tool without experimental or clinical validation, it represents a proof-of-concept for early-stage drug screening rather than a validated clinical aid.
Computational machine learning model development with cross-validation. 2,039 molecular compounds from the MoleculeNet BBBP dataset; no human subjects or clinical populations.. Intervention: Machine learning classification using four algorithms (Logistic Regression, Support Vector Machine, Random Forest, Extreme Gradient Boosting) trained on 15 molecular descriptors. Compared with: Four machine learning algorithms compared against each other; no external benchmark or existing predictive tool comparison reported. n = 2,039.
XGBoost classifier achieved accuracy of 88.97%, precision of 88.92%, recall of 97.76%, F1-score of 93.13%, and ROC-AUC of 0.9282 on the training/validation set Stratified five-fold cross-validation yielded mean ROC-AUC of 0.8982 ± 0.0130, demonstrating model robustness TPSA, HBD, and LogP identified as the three most influential molecular descriptors for BBB permeability prediction via feature importance and SHAP analysis
Safety was not reported in the material analysed. Check the source before drawing any conclusion about harm.
If validated experimentally and clinically, this framework could accelerate early-stage screening of CNS drug candidates by computationally predicting BBB permeability before synthesis. However, the current preprint does not establish clinical utility and should not be used for regulatory decision-making without peer review and independent external validation.
This is an unreviewed preprint presenting a machine learning model for predicting BBB permeability from molecular descriptors; it is methodologically sound but lacks peer review and clinical validation.
As stated by the source record.
Quoted from the source exactly as published.
If validated experimentally and clinically, this framework could accelerate early-stage screening of CNS drug candidates by computationally predicting BBB permeability before synthesis. However, the current preprint does not establish clinical utility and should not be used for regulatory decision-making without peer review and independent external validation.
Graded across the dimensions that decide whether you should act, each from what the source actually supports. There is no single score, and where a dimension was not assessed it says so.
Blood-brain barrier (BBB) permeability is a critical determinant in the development of central nervous system therapeutics because it directly influences the ability of drug candidates to reach their target sites within the brain. In this study, an explainable machine learning framework was developed to predict BBB permeability using molecular descriptors generated from the MoleculeNet BBBP dataset with the RDKit cheminformatics toolkit. Fifteen physicochemical descriptors extracted from 2,039 compounds were used to train four supervised machine learning algorithms, including Logistic Regression, Support Vector Machine (SVM), Random Forest, and Extreme Gradient Boosting (XGBoost). Hyperparameter optimization was performed using GridSearchCV, while model interpretability was investigated using SHapley Additive exPlanations (SHAP). Among the evaluated models, the optimized XGBoost classifier achieved the best predictive performance, with an accuracy of 88.97%, a precision of 88.92%, a recall of 97.76%, an F1-score of 93.13%, and a ROC-AUC of 0.9282. Stratified five-fold cross-validation further demonstrated the robustness of the proposed model, yielding a mean ROC-AUC of 0.8982 +/- 0.0130. Feature importance and SHAP analyses consistently identified TPSA, HBD, and LogP as the most influential molecular descriptors governing BBB permeability prediction. Overall, the proposed framework provides an accurate, interpretable, and computationally efficient approach for BBB permeability prediction and may serve as a valuable tool for the early-stage screening of CNS drug candidates.
Taken from the source record, never inferred. Follow any of these and new work involving them reaches your briefing.