1 - 6 of 6 results


A predictor capable of identifying capsid and tail protein sequences, which are the cornerstones toward viral sequence annotation and viral genome classification. VIRALpro is incorporated into the SCRATCH prediction suite. It uses support vector machines with an accuracy estimated to be between 90% and 97%. Predictions are based on the protein amino acid composition, on the protein predicted secondary structure, as predicted by SSpro, and on a boosted linear combination of HMM e-values obtained from 3,380 HMMs built from multiple sequence alignments of specific fragments - called contact fragments - of both capsid and tail sequences.

VaZyMolO / VirAl enZYme MOdule LOcalization

Defines and classifies viral protein modularity. VaZyMolO enables the handling of viral sequences at the protein level in order to define their modularity. It organizes information about modularity on viral open reading frames from complete genome sequences derived from GenBank and RefSeq. Its annotations are based on a stringent manual checking, specially concerning the boundaries of the modules, and it benefits from virologists’ knowledge.


Identifies the phage proteins located in host cell based on the sequence information. PHPred is a tool to guide related drug discovery. This predictor can become a helpful tool for Pleckstrin Homology (PH) protein analysis and research. Moreover, the feature selection technique can be generalized to other fields of computational biology. In jack-knife cross-validation, this method can discriminate between the bacteriophage proteins located in host cell (PH proteins) and the bacteriophage proteins not located in host cell (non-PH proteins) with maximum overall accuracy of 84.2 per cent.


Predicts phage enzymes and hydrolases at a high discriminative accuracy without having to learn complicated mathematics or programs. PHYPred is a user-friendly webserver that offers a feature selection technique applied to improve the performance. A high-quality benchmark dataset was constructed by setting a series of standards, which can guarantee the reliability of the tool. This method can be used in other fields such as bioinformatics and computational biology and by the vast majority of scholars.

iVIREONS / identification of VIRions by Ensembles Of Neural networkS

A web-based interface to ensembles of trained artificial neural networks that were trained to identify virion structural proteins by voting on translated open reading frames. iVIREONS networks correctly identify, with a high degree of accuracy, ORFs in GenBank that have annotations such as capsid, tape measure, portal,tail, fiber, baseplate, connector, neck, and collar. Additional neural network ensembles have been trained to identify more specific classes of structural proteins, namely major capsid and tail proteins.