Integrated machine learning, molecular docking and ADMET analysis for computational prioritization of S1PR1-targeting drug candidates

Sphingosine-1-phosphate receptor 1 (S1PR1) is a G protein-coupled receptor involved in immune-cell trafficking and has attracted considerable interest as a pharmacological target. The availability of structurally characterized ligand-bound S1PR1 provides an opportunity to combine ligand-based machine learning with structure-based screening to identify additional compounds with potential S1PR1 binding activity. This study aimed to develop and evaluate an integrated computational workflow to identify and prioritize compounds predicted to interact with S1PR1 from an external drug library. Specifically, we investigated whether molecular fingerprints and physicochemical descriptors could reliably predict S1PR1 activity and whether sequential machine-learning screening, drug-likeness filtering, molecular docking, interaction analysis, and absorption, distribution, metabolism, excretion, and toxicity (ADMET) prediction could prioritize candidates for further investigation. A curated ChEMBL dataset containing 3,300 unique compounds was used to develop random forest, extreme gradient boosting (XGBoost), logistic regression, and support vector machine models using scaffold-separated training and independent test sets. Random forest and XGBoost demonstrated strong predictive discrimination, with independent-test area under the receiver operating characteristic curve values of 0.9198 and 0.9252, respectively. The optimized random forest model was applied to 12,289 DrugBank compounds, of which 277 met the predefined prediction threshold. Subsequent Lipinski filtering retained 56 compounds for docking, with 54 yielding successful docking results. Integration of predicted activity, docking scores, physicochemical properties, protein–ligand interactions, and ADMET profiles prioritized five compounds: udifitimod, AZD8529, tedizolid phosphate, mocravimod, and BMS-986104. Chemical-space analysis indicated that many DrugBank compounds were structurally distant from the model-development chemical space. Overall, the workflow provides a computational strategy for prioritizing candidates for S1PR1-binding compounds for subsequent experimental evaluation.

Authors

Publication Details

Journal
Innovative Medicines & Omics
Published
2026-11-17
DOI
https://doi.org/10.36922/imo026300055
Primary Topic
Computational Drug Discovery Methods
Type
article
Field-Weighted Citation Impact
0.00
Controls
|||
ALL TIME
JAN
FEB
MAR
APR
MAY
JUN
JUL
AUG
SEP
OCT
NOV
article

Integrated machine learning, molecular docking and ADMET analysis for computational prioritization of S1PR1-targeting drug candidates

Elamathi Natarajan, Akshata P. Pauskar
3 links
Innovative Medicines & Omics
Computational Drug Discovery Methods
article

Integrated machine learning, molecular docking and ADMET analysis for computational prioritization of S1PR1-targeting drug candidates

Elamathi Natarajan, Akshata P. Pauskar
article en

Abstract

Sphingosine-1-phosphate receptor 1 (S1PR1) is a G protein-coupled receptor involved in immune-cell trafficking and has attracted considerable interest as a pharmacological target. The availability of structurally characterized ligand-bound S1PR1 provides an opportunity to combine ligand-based machine learning with structure-based screening to identify additional compounds with potential S1PR1 binding activity. This study aimed to develop and evaluate an integrated computational workflow to identify and prioritize compounds predicted to interact with S1PR1 from an external drug library. Specifically, we investigated whether molecular fingerprints and physicochemical descriptors could reliably predict S1PR1 activity and whether sequential machine-learning screening, drug-likeness filtering, molecular docking, interaction analysis, and absorption, distribution, metabolism, excretion, and toxicity (ADMET) prediction could prioritize candidates for further investigation. A curated ChEMBL dataset containing 3,300 unique compounds was used to develop random forest, extreme gradient boosting (XGBoost), logistic regression, and support vector machine models using scaffold-separated training and independent test sets. Random forest and XGBoost demonstrated strong predictive discrimination, with independent-test area under the receiver operating characteristic curve values of 0.9198 and 0.9252, respectively. The optimized random forest model was applied to 12,289 DrugBank compounds, of which 277 met the predefined prediction threshold. Subsequent Lipinski filtering retained 56 compounds for docking, with 54 yielding successful docking results. Integration of predicted activity, docking scores, physicochemical properties, protein–ligand interactions, and ADMET profiles prioritized five compounds: udifitimod, AZD8529, tedizolid phosphate, mocravimod, and BMS-986104. Chemical-space analysis indicated that many DrugBank compounds were structurally distant from the model-development chemical space. Overall, the workflow provides a computational strategy for prioritizing candidates for S1PR1-binding compounds for subsequent experimental evaluation.

Innovative Medicines & OmicsVol. 0(0)
Good health and well-being
Computational Drug Discovery Methods
AI Navigator

Ask Laika to Summarize, Analyze, and Connect papers live on the map.

Summarize Papers & Methodologies

Extract key findings, datasets, and comparative methods across publications.

Benchmark Rankings & Visual Analytics

Rank top research institutions, authors, funders, topics, and journals by Field-Weighted Citation Impact (FWCI) and paper volume with instant charts.

Connect Distant Disciplines

Bridge topological clusters on the map to find hidden collaborative intersections.

Laika access coming soon