Developing of tool for extracting protein descriptors of position specific scoring matrix

Abstract: Feature extraction or feature encoding is a fundamental step in the construction of high-quality machine learning-based models. Specifically, this step is key to determining the effectiveness of trained models in bioinformatics applications.[1] In the last two decades, a variety of feature encoding schemes have been proposed in order to exploit useful patterns from protein sequences. Such schemes are often based on sequence information or representation of physicochemical properties. [2] Although direct features derived from sequences themselves (such as amino acid compositions, dipeptide compositions and counting of k-mers) are regarded as essential for training models, an increasing number of studies have shown that evolutionary information in the form of PSSM profiles is much more informative than sequence information alone. there is no comprehensive, simple tool for extracting all of features from the PSSM matrix and displaying it in the output. In this study, the goal is to develop a comprehensive tool that can be used as an input to the protein sequence and produce PSSM matrix output with all of these descriptors

Keywords:

position specific scoring matrix , protein descriptors , feature extraction , machine learning , based models