TemStaPro Datasets

This dataset contains protein sequences used to train, validate, and test binary classifiers that form TemStaPro program, which is applied for protein thermostability prediction with respect to 6 temperature thresholds from 40 to 65 degrees Celsius using a step of 5 degrees. The data is given in files of FASTA format. Each protein sequence has a header made of 3 values separated by vertical bar symbols: organism's, to which the protein belongs, UniParc taxonomy identifier; UniProtKB/TrEMBL identifier of the protein sequence; organism's growth temperature taken from the dataset of growth temperatures of over 21 thousand organisms (Engqvist, 2018). TemStaPro-Major set is composed of 9 files: training, validation, imbalanced and balanced for each of 6 binary classifiers' thresholds testing sets. TemStaPro-Minor set is composed of training, validation, and testing files all balanced for 65 degrees Celsius temperature threshold.<br> <br> The detailed description is given in the corresponding paper (https://doi.org/10.1101/2023.03.27.534365). If you use the data from this dataset, please cite both the paper and the DOI of the dataset.