DevInD (Development of Industrial Drug molecules) is a specialized computational platform designed to provide comprehensive information and predictive models for drug molecules currently used or developed in the pharmaceutical industry. This resource serves as a centralized repository to assist researchers in understanding the chemical and biological properties of successful industrial drug candidates.
Web Server: https://webs.iiitd.edu.in/raghava/denvind/
Vivek Dhar Dwivedi, Aditya Arya, Pardeep Yadav, Rajesh Kumar, Vinod Kumar, Gajendra P S Raghava, DenvInD: dengue virus inhibitors database for clinical and molecular research, Briefings in Bioinformatics, Volume 22, Issue 3, May 2021, https://doi.org/10.1093/bib/bbaa098
This dataset can also be found on Zenodo at https://doi.org/10.5281/zenodo.20094661
The pharmaceutical industry follows rigorous standards for selecting molecules with optimal pharmacokinetic and pharmacodynamic profiles. DevInD consolidates data on these "industrial-strength" molecules, which often differ significantly from common chemical libraries or early-stage leads.
- Data Compilation: The platform is built upon a manually curated dataset of drug molecules that have reached various stages of industrial development and clinical trials.
- Focus on Drug-Likeness: It emphasizes the structural features and molecular descriptors that contribute to a molecule's success in an industrial setting.
- Curated Molecules: Contains detailed profiles of molecules, including their chemical structures, target proteins, and therapeutic categories.
-
Pharmacological Data: Provides information on
$IC_{50}$ ,$EC_{50}$ , and other potency measures relevant to industrial drug standards.
- Industrial Potency Prediction: Utilizing machine learning models (such as SVM and Random Forest), the platform allows users to predict the likelihood of a query molecule being an effective industrial drug candidate.
- Descriptor Analysis: Computes a wide array of 1D, 2D, and 3D molecular descriptors to characterize the chemical space of the molecules.
- Search and Browse: Users can easily query the database by drug name, chemical scaffold, or therapeutic use.
- Similarity Search: Tools to find industrial molecules that are structurally similar to a user's query compound.
- Property Visualization: Interactive tools to visualize the distribution of physicochemical properties across the industrial drug dataset.
- Drug Repurposing: Identifying existing industrial molecules that may have therapeutic potential for new disease indications.
- Lead Prioritization: Helping researchers prioritize lead compounds by comparing them to the structural profiles of established industrial drugs.
- Pharmacoinformatics: Providing a high-quality dataset for training new AI/ML models in the field of drug discovery and development.
Prof. Gajendra P. S. Raghava (Corresponding Author)
Department of Computational Biology, Indraprastha Institute of Information Technology (IIIT Delhi), New Delhi, India.
DevInD was developed with support from the Department of Biotechnology (DBT) and the Council of Scientific and Industrial Research (CSIR), Government of India. Infrastructure and facilities were provided by IIIT-Delhi.