FireProtDB: database of manually curated protein stability data

Abstract

The majority of naturally occurring proteins have evolved to function under mild conditions inside the living organisms. One of the critical obstacles for the use of proteins in biotechnological applications is their insufficient stability at elevated temperatures or in the presence of salts. Since experimental screening for stabilizing mutations is typically laborious and expensive, in silico predictors are often used for narrowing down the mutational landscape. The recent advances in machine learning and artificial intelligence further facilitate the development of such computational tools. However, the accuracy of these predictors strongly depends on the quality and amount of data used for training and testing, which have often been reported as the current bottleneck of the approach. To address this problem, we present a novel database of experimental thermostability data for single-point mutants FireProtDB. The database combines the published datasets, data extracted manually from the recent literature, and the data collected in our laboratory. Its user interface is designed to facilitate both types of the expected use: (i) the interactive explorations of individual entries on the level of a protein or mutation and (ii) the construction of highly customized and machine learning-friendly datasets using advanced searching and filtering. The database is freely available at https://loschmidt.chemi.muni.cz/fireprotdb.

Towards Compilation of Balanced Protein Stability Datasets: Flattening the ΔΔG Curve through Systematic Under-sampling

“…Furthermore, automated and flexible implementation of this approach ensure one to easily adapt it to tailor the stability subset of need. One with a much larger parent datasets such as ThermoMutDB, 25 FireProtDB 62 and ProtaBank 63 can apply a more rigorous constraint selection than ours with the toy dataset PON-tstab. Different reduced alphabets can be implemented and further the sampling ΔΔG interval can be adjusted such that attenuation of this value would enforce a more stringent elimination within mutation groups and thus result in a larger subset.…”

Applications of AlphaFold beyond Protein Structure Prediction

“…In this study, we used engineered and designed sequences to investigate the applicability of AlphaFold to problems other than structure prediction for naturally occurring sequences. We used mainly two types of sequences: point mutations with experimentally measured stability changes 18 , and sequences designed to fold to target protein structures using a modified algorithm based on ProDCoNN 11 . We found that the representations learned by AlphaFold during the prediction process can accurately predict the stability changes of point mutations.…”

“…To study the correlation between AlphaFold confidence scores and the experimentally measured stability changes, we randomly selected 3507 experiments from protein single-point mutants stability database FireProtDB 13 , corresponding to 1251 mutants from 86 protein chains. The dataset contains 2557 experiments with Gibbs free energy changes (ΔΔG) upon mutation and 952 experiments with changes in melting temperatures (ΔTm).…”

Fundamentals to function: Quantitative and scalable approaches for measuring protein stability

“…As new technologies generate more stability data, there is a rising need to systematically combine data across different methods, experiments, and conditions. Three databases-FireProtDB (Stourac et al, 2021), ProtaBank (Wang et al, 2018a(Wang et al, , 2018b, and ProTherm (Gromiha et al, 1999)-have been created to collect, organize, and validate stability data from the literature. This growing collection of data also underscores the importance of establishing best practices for performing and reporting stability measurements, as has been pioneered for other types of complex data from diverse sources (Conesa et al, 2016;Jarmoskaite et al, 2020).…”