FireProtDB

FireProtDB provides a curated repository of experimental thermostability data for single-point protein mutants to support development, training, and evaluation of computational predictors and to inform design of stabilizing mutations for biotechnological applications under conditions such as elevated temperature and high-salt concentration.

Key Features:

• Experimental thermostability data: Contains curated measurements of thermostability for single-point protein mutants.
• Data aggregation: Amalgamates published datasets, manually extracted recent literature data, and proprietary laboratory-collected data.
• Customized dataset construction: Provides advanced searching and filtering to create tailored datasets optimized for machine learning applications.
• Support for in silico predictors: Supplies high-quality experimental data to improve training and testing of computational stability prediction tools.

Scientific Applications:

• Machine learning model development: Training and benchmarking of machine learning and AI-based protein stability predictors.
• Rational protein design: Informing mutational strategies to increase protein stability for industrial conditions such as elevated temperature and high-salt concentration.
• Benchmarking and validation: Providing experimental ground truth for evaluation and validation of computational prediction methods.

Methodology:

Aggregates data from published datasets, manually extracts stability data from recent literature, and incorporates proprietary laboratory-collected measurements, and enables dataset construction via advanced searching and filtering.