What is DRLiPS?
DRLiPS is a web server for the prediction of druggable binding pockets for small molecules, in an input RNA structure. Additionally, the druggability score of a user-defined pocket in all models of an NMR structure can also be computed. DRLiPS is powered by a machine learning model which computes RNA structure-based features and scores binding pockets with a value between 0 and 1, with higher scores indicating high druggability.
What is the dataset used in the DRLiPS model?
The dataset of RNA-small molecule binding pockets for model development was curated from the RCSB PDB database and the HARIBOSS database. The negative dataset was curated based on three strategies namely: all-to-all blind docking, backbone motif matching and in silico pocket prediction. For more information on the dataset, please visit the Statistics page of the web server.
How to define a known binding pocket for druggability prediction?
The binding pocket must be provided as a comma-separated list of residues/nucleotides, along with their residue name and chain ID. Examples of the input format expected by DRLiPS is explained in the Tutorials page.
Can I use DRLiPS even if I don't know any binding pockets in my RNA structure?
Yes, DRLiPS can be used without defining any binding pockets of interest. In such cases, DRLiPS will use the fpocket prediction program in the background and provide the predicted druggability scores for each predicted binding pocket as the result.
Can DRLiPS work with apo RNA structures and modelled structures?
Yes, DRLiPS can generalize well to apo and modelled RNA structures, even though it was trained only with holo RNA structures. However, if the backbone RMSD between the apo and holo conformations of the RNA exceeds 2.5 Å, the model predictions for apo state can greatly differ from that of the holo state.
Can DRLiPS predict the effect of binding site mutations on the RNA druggability?
Yes, DRLiPS can capture the effect of even single point mutations. The change in DRLiPS score for the mutated structure has been shown indicate the corresponding increase/decrease in promiscuity and druggability of a binding site. Check out some of the interesting case studies in this direction available in the associated article!
Can DRLiPS perform batch calculations?
No, currently DRLiPS can currently handle only one RNA structure at a time. However, batch calculation capabilities will be implemented in future in the server, based on user feedback.
What is the browser compatibility of DRLiPS?
DRLiPS is powered by HTML5, CSS, PHP, Python3 and BootstrapJS. It is best viewed in Firefox, but is also compatible with Chrome, IE10 and Opera browsers.
Can DRLiPS be used for commercial purposes?
No, DRLiPS is strictly released for academic and non-profit research purposes only. For more details on the terms of use, please visit the About Us page.
© 2024, Protein Bioinformatics Lab, Indian Institute of Technology Madras All Rights Reserved. This server is maintained by Department of Biotechnology (IIT-M) |
![]() |