RNA Ligand-Based Similarity Analysis

Note

This repository contains data and scripts for analyzing ligand-based similarity metrics and molecular fingerprints in the context of RNA-targeted small molecule discovery. The content is related to scientific publication: Bajusz, D., Rácz, A., Bujnicki, J. M., & Stefaniak, F. (2025). Evaluation of single-template ligand-based methods for the discovery of small-molecule nucleic acid binders. Briefings in Bioinformatics, 26(6), bbaf620, https://doi.org/10.1093/bib/bbaf620

DOI: 10.5281/zenodo.15683151

Repository Structure

📂 1-fingerprints/

Contains molecular fingerprint datasets in compressed CSV format. Each file represents a different fingerprinting method (e.g., CDK ECFP, FCFP, MACCS, PubChem, Indigo, MAP4, OpenBabel, RDKit, random bits).

📂 descriptors/

Contains molecular descriptors calculated for all ligands.

📂 3-metrics/

Contains statistical analyses and metrics, including t-tests comparing the performance of different methods and random baselines.

• metrics__ttest.csv: Summary of statistical tests for various methods.
• combined_best_metrics.csv.gz: combined values for all methods and all similarity metrics
• individual_metrics/: Individual metrics for fingerprints (all subdirectories) and individual ligand-based methods (tanimoto directory). The best-performing variants of all methods are in _best directory.

📂 p-values

Contains average values for the performance metrics with the statistical significance p-values

📂 4-ANOVA/

Contains results of ANOVA analyses and interaction means for different factors:

• F1_weighted_means.csv, F2_weighted_means.csv, F3_weighted_means.csv: Weighted means for different factors.
• F1_F2_interaction_means.csv, F2_F3_interaction_means.csv, F3_F1_interaction_means.csv: Interaction means between factors.

📂 Pooled_ROC

Plots of pooled ROC curves for selected methods and all targets

📂 5-case-study/

Contains case studies data for Glutamine_RS.

📂 6-FP3_analysis/

Contains results of FP3 fingerprint analysis.

📂 7-docking/

Results of molecular docking:

• 1-macromolecules - nucleic acid targets used for docking
• 1-results-sdf/ - docked poses
• 2-scores/ - scores extracted from sdf files
• 3-performance/ - performance metrics

📂 code/

Programs used to perform calculations.

📰 Citation

If you use this repository or its data in your research, please cite the relevant publication or acknowledge the authors:

Bajusz, D., Rácz, A., Bujnicki, J. M., & Stefaniak, F. (2025). Evaluation of single-template ligand-based methods for the discovery of small-molecule nucleic acid binders. Briefings in Bioinformatics, 26(6), bbaf620, https://doi.org/10.1093/bib/bbaf620