TY - JOUR
T1 - Automatic segmentation of Potyviridae family polyproteins
AU - Vargas, Jheyson Faride
AU - Velasco, Jairo Andrés
AU - Alvarez, Gloria Inés
AU - Linares, Diego Luis
AU - Bravo, Enrique
N1 - Publisher Copyright:
© 2015 Inderscience Enterprises Ltd.
PY - 2015
Y1 - 2015
N2 - We describe an automatic segmentation method for polyproteins of the viruses belonging to the Potyviridae family. It uses machine learning techniques in order to predict the cleavage site which define the segments in which said polyproteins are cut in their process of functional maturation. The segmentation application is publicly available for use on a website and it can be accessed through the web service interface too. The prediction models have an average sensitivity of 0.79 and a Matthews correlation coefficient average of 0.23. This method is capable of predicting correctly (coinciding with previously published segmentation) the segmentation of sequences which come from Potyvirus and Rymovirus, genera. However accurate prediction capabilities are affected when faced with either atypical sequences or viruses belonging to less common genera in the Potyviridae family. Future work will focus on establishing greater flexibility in this sense.
AB - We describe an automatic segmentation method for polyproteins of the viruses belonging to the Potyviridae family. It uses machine learning techniques in order to predict the cleavage site which define the segments in which said polyproteins are cut in their process of functional maturation. The segmentation application is publicly available for use on a website and it can be accessed through the web service interface too. The prediction models have an average sensitivity of 0.79 and a Matthews correlation coefficient average of 0.23. This method is capable of predicting correctly (coinciding with previously published segmentation) the segmentation of sequences which come from Potyvirus and Rymovirus, genera. However accurate prediction capabilities are affected when faced with either atypical sequences or viruses belonging to less common genera in the Potyviridae family. Future work will focus on establishing greater flexibility in this sense.
KW - Artificial neural networks
KW - Automatic segmentation
KW - Hidden Markov models
KW - Polyprotein
KW - Potyviridae
KW - Potyvirus
UR - http://www.scopus.com/inward/record.url?scp=84948756577&partnerID=8YFLogxK
U2 - 10.1504/IJBRA.2015.073238
DO - 10.1504/IJBRA.2015.073238
M3 - Article
AN - SCOPUS:84948756577
SN - 1744-5485
VL - 11
SP - 525
EP - 539
JO - International Journal of Bioinformatics Research and Applications
JF - International Journal of Bioinformatics Research and Applications
IS - 6
ER -