Prediction of glutathionylation sites in proteins using minimal sequence information and their experimental validation.

摘要: S-glutathionylation of proteins plays an important role in various biological processes and is known to be protective modification during oxidative stress. Since, experimental detection labor intensive time consuming, bioinformatics based approach a viable alternative. Available methods require relatively longer sequence information, which may prevent prediction if information incomplete. Here, we present model predict glutathionylation sites from pentapeptide sequences. It upon differential association amino acids with glutathionylated non-glutathionylated cysteines database experimentally verified This data was used calculate position dependent F-scores, measure how particular acid at affect the likelihood event. Glutathionylation-score (G-score), indicating propensity undergo glutathionylation, calculated using position-dependent F-scores for each amino-acid. Cut-off values were prediction. Our returned accuracy 58% Matthew's correlation-coefficient (MCC) value 0.165. On independent dataset, our outperformed currently available model, spite needing much less information. Pentapeptide motifs having high abundance among identified. A list potential hotspot sequences obtained by assigning G-scores subsequent Protein-BLAST analysis revealed total 254 putative glutathionable proteins, number already glutathionylated. predicted 93.93% proteins. Outcome this study assist discovering novel finding candidate glutathionylation.