TmProt: A Web Server for Predicting Protein Melting Temperatures from Sequences
Karen Pailozian, Jiri Damborsky, Stanislav Mazurenko. Manuscript in preparation, 2024.

TmProt is a web tool for sequence-based prediction of protein melting temperatures. Despite recent progress in high-throughput protein characterization, protein databases remain largely unannotated, presenting ample opportunities for discovering promising biocatalysts using automated tools like EnzymeMiner.

Knowing the melting temperature (T_m) of a protein is valuable in this respect, as protein stability is a critical property for many biotechnological and biopharmaceutical applications. However, the complexity of the mechanisms governing protein stability makes its prediction challenging. To tackle this challenge, we have collected and manually curated a large dataset with protein melting temperatures. This dataset has enabled us to train a state-of-the-art ML-based predictor.

After pasting a set of sequences in the FASTA format, the web server will return predicted melting temperatures and their confidence scores. We are currently at the last stages of validating the model and implementing it as a web server designed to make the prediction easy to understand for the user. The release is planned for the end of 2023.

Software tools developed by the Loschmidt Laboratories.