Accurate prediction of protein enzymatic class by N-to-1 Neural Networks

Hdl Handle:
http://hdl.handle.net/10147/265280
Title:
Accurate prediction of protein enzymatic class by N-to-1 Neural Networks
Authors:
Volpato, Viola; Adelfio, Alessandro; Pollastri, Gianluca
Citation:
BMC Bioinformatics. 2013 Jan 14;14(Suppl 1):S11
Issue Date:
14-Jan-2013
URI:
http://dx.doi.org/10.1186/1471-2105-14-S1-S11; http://hdl.handle.net/10147/265280
Abstract:
Abstract We present a novel ab initio predictor of protein enzymatic class. The predictor can classify proteins, solely based on their sequences, into one of six classes extracted from the enzyme commission (EC) classification scheme and is trained on a large, curated database of over 6,000 non-redundant proteins which we have assembled in this work. The predictor is powered by an ensemble of N-to-1 Neural Network, a novel architecture which we have recently developed. N-to-1 Neural Networks operate on the full sequence and not on predefined features. All motifs of a predefined length (31 residues in this work) are considered and are compressed by an N-to-1 Neural Network into a feature vector which is automatically determined during training. We test our predictor in 10-fold cross-validation and obtain state of the art results, with a 96% correct classification and 86% generalized correlation. All six classes are predicted with a specificity of at least 80% and false positive rates never exceeding 7%. We are currently investigating enhanced input encoding schemes which include structural information, and are analyzing trained networks to mine motifs that are most informative for the prediction, hence, likely, functionally relevant.
Item Type:
Journal Article

Full metadata record

DC FieldValue Language
dc.contributor.authorVolpato, Viola-
dc.contributor.authorAdelfio, Alessandro-
dc.contributor.authorPollastri, Gianluca-
dc.date.accessioned2013-01-14T14:07:28Z-
dc.date.available2013-01-14T14:07:28Z-
dc.date.issued2013-01-14-
dc.identifier.citationBMC Bioinformatics. 2013 Jan 14;14(Suppl 1):S11-
dc.identifier.urihttp://dx.doi.org/10.1186/1471-2105-14-S1-S11-
dc.identifier.urihttp://hdl.handle.net/10147/265280-
dc.description.abstractAbstract We present a novel ab initio predictor of protein enzymatic class. The predictor can classify proteins, solely based on their sequences, into one of six classes extracted from the enzyme commission (EC) classification scheme and is trained on a large, curated database of over 6,000 non-redundant proteins which we have assembled in this work. The predictor is powered by an ensemble of N-to-1 Neural Network, a novel architecture which we have recently developed. N-to-1 Neural Networks operate on the full sequence and not on predefined features. All motifs of a predefined length (31 residues in this work) are considered and are compressed by an N-to-1 Neural Network into a feature vector which is automatically determined during training. We test our predictor in 10-fold cross-validation and obtain state of the art results, with a 96% correct classification and 86% generalized correlation. All six classes are predicted with a specificity of at least 80% and false positive rates never exceeding 7%. We are currently investigating enhanced input encoding schemes which include structural information, and are analyzing trained networks to mine motifs that are most informative for the prediction, hence, likely, functionally relevant.-
dc.titleAccurate prediction of protein enzymatic class by N-to-1 Neural Networks-
dc.typeJournal Article-
dc.language.rfc3066en-
dc.rights.holderViola Volpato et al.; licensee BioMed Central Ltd.-
dc.description.statusPeer Reviewed-
dc.date.updated2013-01-14T12:04:54Z-
All Items in Lenus, The Irish Health Repository are protected by copyright, with all rights reserved, unless otherwise indicated.