TY - JOUR
T1 - Improving succinylation prediction accuracy by incorporating the secondary structure via helix, strand and coil, and evolutionary information from profile bigrams
AU - Dehzangi, Abdollah
AU - López, Yosvany
AU - Lal, Sunil Pranit
AU - Taherzadeh, Ghazaleh
AU - Sattar, Abdul
AU - Tsunoda, Tatsuhiko
AU - Sharma, Alok
N1 - Publisher Copyright:
© 2018 Dehzangi et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
PY - 2018/2
Y1 - 2018/2
N2 - Post-translational modification refers to the biological mechanism involved in the enzymatic modification of proteins after being translated in the ribosome. This mechanism comprises a wide range of structural modifications, which bring dramatic variations to the biological function of proteins. One of the recently discovered modifications is succinylation. Although succinylation can be detected through mass spectrometry, its current experimental detection turns out to be a timely process unable to meet the exponential growth of sequenced proteins. Therefore, the implementation of fast and accurate computational methods has emerged as a feasible solution. This paper proposes a novel classification approach, which effectively incorporates the secondary structure and evolutionary information of proteins through profile bigrams for succinylation prediction. The proposed predictor, abbreviated as SSEvol-Suc, made use of the above features for training an AdaBoost classifier and consequently predicting succinylated lysine residues. When SSEvol-Suc was compared with four benchmark predictors, it outperformed them in metrics such as sensitivity (0.909), accuracy (0.875) and Matthews correlation coefficient (0.75).
AB - Post-translational modification refers to the biological mechanism involved in the enzymatic modification of proteins after being translated in the ribosome. This mechanism comprises a wide range of structural modifications, which bring dramatic variations to the biological function of proteins. One of the recently discovered modifications is succinylation. Although succinylation can be detected through mass spectrometry, its current experimental detection turns out to be a timely process unable to meet the exponential growth of sequenced proteins. Therefore, the implementation of fast and accurate computational methods has emerged as a feasible solution. This paper proposes a novel classification approach, which effectively incorporates the secondary structure and evolutionary information of proteins through profile bigrams for succinylation prediction. The proposed predictor, abbreviated as SSEvol-Suc, made use of the above features for training an AdaBoost classifier and consequently predicting succinylated lysine residues. When SSEvol-Suc was compared with four benchmark predictors, it outperformed them in metrics such as sensitivity (0.909), accuracy (0.875) and Matthews correlation coefficient (0.75).
UR - http://www.scopus.com/inward/record.url?scp=85042089700&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=85042089700&partnerID=8YFLogxK
U2 - 10.1371/journal.pone.0191900
DO - 10.1371/journal.pone.0191900
M3 - Article
C2 - 29432431
AN - SCOPUS:85042089700
VL - 13
JO - PLoS One
JF - PLoS One
SN - 1932-6203
IS - 2
M1 - e0191900
ER -