Abstract
Background: Unstructured regions in proteins can vary from several amino acid residues to a completely disordered sequence. Since such regions play an important role in the protein functioning, much attention is being paid to their prediction. Special different programs are available for this purpose; however, predictions obtained vary from protein to protein.
Objective: In this article, our motivation is the investigation of the high prediction accuracy of flexible loops in G-proteins family with FoldUnfold program due to crucial functions associated with these regions.
Method: For prediction of loops in the G-proteins we used programs as RONN, DisEMBL, Glob- Plot2, IUPred, PONDR, FoldIndex, MobiDB and FoldUnfold. As a criterion of reliability of predicting disordered regions, we have chosen comparison with the regions known from the 3D structures. Collection, data analysis and statistical analysis were performed using Python 3.3. and R version 3.2.0.
Results: For 23 G-proteins, the FoldUnfold program predicts loops with the average precision of 60-80%. It is seen that our program enables better prediction of loop positions than other programs. Statistically significant weak negative correlation exists between the average number of closed residues according to the FoldUnfold program and the Debye-Waller factors. Investigations of the G-proteins with the posttranslational modifications revealed additional flexible properties of the residues involved in the attachment of fatty acids.
Conclusion: Our research demonstrates additional possibilities and the high prediction accuracy of the FoldUnfold program for prediction of flexible regions and characteristics of individual residues in different protein family.
Keywords: G-proteins, FoldUnfold program, flexible regions, functional loops, Debye-Waller factor, posttranslational modifications.
Graphical Abstract
Related Journals
Current Protein & Peptide Science
Related Books
Frontiers in Protein and Peptide Sciences
17
5