Extensive comparison of protein sequence-based bioinformatics applications for predicting lysine succinylation sites: a comparative review
| Autor: | Hussam Alsharif |
|---|---|
| Jazyk: | angličtina |
| Rok vydání: | 2024 |
| Předmět: | |
| Zdroj: | Biotechnology & Biotechnological Equipment, Vol 38, Iss 1 (2024) |
| Druh dokumentu: | article |
| ISSN: | 13102818 1314-3530 1310-2818 |
| DOI: | 10.1080/13102818.2024.2425694 |
| Popis: | Lysine succinylation is a post-translational modification that occurs when a succinyl group bonds with a lysine residue and changes the polarity of lysine from positive to negative, causing significant changes in the structure and functioning of proteins. Lysine succinylation occurs on various proteins and is crucial in the cellular and biological processes of eukaryotic and prokaryotic organisms. Succinylation site identification is an area of high research interest, and sequence-based prediction methods using machine learning and deep learning have been developed based on experimentally confirmed data of succinylation sites, aiming to be highly accurate, robust, quick, and cost-efficient. However, despite the usefulness of these methods, different issues must be addressed when building a model of lysine succinylation. As succinylation site predictors become more abundant, it is crucial to assess their advantages and drawbacks to identify potential issues and improve the efficiency of predicting succinylation sites. Among the multiple studies that have employed machine-learning and deep-learning applications, few have systematically examined computational issues. Hence, in this review, we summarize the challenges and restrictions in the development of succinylation prediction models and provide guidelines for more suitable and efficient computational methods. |
| Databáze: | Directory of Open Access Journals |
| Externí odkaz: |
|