TL;DR
PROTAC-DB is a new, comprehensive, web-based database designed to accelerate the rational design of Proteolysis-Targeting Chimeras (PROTACs). It consolidates structural information, biological activities, and physicochemical properties of PROTACs, their warheads, E3 ligands, and linkers, offering powerful search and filtering tools for researchers.
What Problem Does This Paper Address?
Proteolysis-Targeting Chimeras (PROTACs) represent a revolutionary therapeutic technology that hijacks the cell's natural protein degradation machinery to eliminate disease-causing proteins. Unlike traditional drugs that inhibit protein function, PROTACs induce targeted degradation, offering potential advantages for "undruggable" targets. However, the design of effective PROTACs is complex, involving the precise assembly of three distinct components: a warhead that binds the target protein, an E3 ligase ligand that recruits an E3 ligase, and a linker connecting them. This complexity, coupled with the need for specific physicochemical and biological properties, makes rational design a significant challenge. Existing databases offer limited information on PROTACs, hindering researchers' ability to explore structure-activity relationships and optimize PROTAC design.
Methods
The PROTAC-DB database was constructed through a systematic data collection and processing pipeline.
- Literature Search: Information was gathered by searching PubMed using keywords like "degrader*" or "PROTAC." The focus was on small-molecule PROTACs, excluding peptide-based and HaloTag PROTACs.
- Data Extraction: Chemical structures and biological activities (degradation capacities, binding affinities, and cellular activities) were manually extracted from the literature.
- Data Organization: PROTACs were systematically broken down into their constituent parts: warheads, E3 ligands, and linkers. This modular approach allows for the analysis of individual components and their contributions to PROTAC function.
- Physicochemical Property Calculation: Ten key physicochemical properties relevant to drug-likeness were calculated using the RDKit toolkit and ALOGPS.
- Database Development: The database was built using the Python web framework Tornado and deployed on a Linux server. PostgreSQL was used as the database management system. OpenEye Python toolkits were employed for visualizing 2D chemical structures, and ChemDoodle was integrated as a molecule editor for user-defined queries.
Datasets
PROTAC-DB itself is a curated dataset. It currently contains:
- 1662 PROTACs
- 202 warheads
- 65 E3 ligands
- 806 linkers
The data was collected from scientific literature and supplemented with information from existing databases such as PubChem, ChEMBL, and BindingDB.
Key Insights and Ideas
- Modular Design is Key: The database's structure, separating PROTACs into warheads, E3 ligands, and linkers, highlights the modular nature of PROTAC design and facilitates the analysis of each component's role.
- Beyond Lipinski's Rules: PROTACs often have higher molecular weights than traditional drugs, potentially violating Lipinski's "rule of five." PROTAC-DB's inclusion of physicochemical properties allows researchers to explore these deviations and understand their implications for drug-likeness.
- Comprehensive Data Integration: PROTAC-DB goes beyond simple structure storage by integrating diverse data types, including degradation capacities (DC50, Dmax), binding affinities (Kd, Ki, IC50), and cellular activities (IC50, EC50), providing a holistic view of PROTAC performance.
- User-Friendly Interface: The web-based interface with text-based and structure-based search options, along with filtering tools, significantly lowers the barrier to entry for researchers wanting to explore PROTAC data.
Key Results
PROTAC-DB currently houses data on 1662 PROTACs, 202 warheads, 65 E3 ligands, and 806 linkers. The database offers:
- Dual Search Capabilities: Users can search by target name, compound name, or ID (text-based) or by inputting a SMILES string, uploading a file, or sketching a molecule (structure-based).
- Detailed Information Pages: Each PROTAC, warhead, and E3 ligand has a dedicated page displaying its structure, associated data, and links to similar compounds.
- Activity Data Visualization: Degradation capacities, binding affinities, and cellular activities are presented in datasheets and detailed information pages, with Western blot figures included for PROTACs to visually demonstrate degradation.
- Physicochemical Property Filtering: A filtering tool allows users to refine search results based on properties like molecular weight, log P, and solubility.
- Data Download: All collected data is available for download in SDF or CSV formats.
Limitations and Future Work
The authors acknowledge that the technology is still maturing, and the design of PROTACs remains a challenge. While PROTAC-DB aims to address this by providing a comprehensive resource, the database is a living entity. The authors plan to:
- Continuously Add New Data: As new PROTACs are discovered and published, the database will be updated.
- Improve Usability: Ongoing efforts will be made to enhance the web interface and user experience.
The database is freely accessible at http://cadd.zju.edu.cn/protacdb/.
Summary updated