|
|
MSLVP is Support Vector Machine (SVM) based two-tier prediction algorithm for annotation of viral protein subcellular localization
First-Tier predicts whether input sequence belong to single/double/multiple location
Second tier predicts specific eight, four and six sub-categories within first-tier locations
Algorithm is developed using comprehensive data set at 90% and 30% sequence identity.
Protein sequence features like amino acid composition (AAC), dipeptide composition (DPC), physicochemical properties (PHY) and their hybrids were employed
"one-versus-other" and "one-versus-one" classification approaches were applied
Average accuracy for hybrid model is about 90-99% for training and on independent data set respectively
It is a comprehensive predictor to annotate subcellular localization of viral proteins with higher accuracy
It will be helpful in understanding the functional annotation of viral proteins and potential drug targets
|
