Show simple item record

dc.contributor.authorFaraggi, Eshel
dc.contributor.authorXue, Bin
dc.contributor.authorZhou, Yaoqi
dc.date.accessioned2017-05-03T15:56:04Z
dc.date.available2017-05-03T15:56:04Z
dc.date.issued2009
dc.date.modified2014-05-28T22:28:01Z
dc.identifier.issn0887-3585
dc.identifier.doi10.1002/prot.22193
dc.identifier.urihttp://hdl.handle.net/10072/57473
dc.description.abstractThis article attempts to increase the prediction accuracy of residue solvent accessibility and real-value backbone torsion angles of proteins through improved learning. Most methods developed for improving the back propagation algorithm of artificial neural networks are limited to small neural networks. Here, we introduce a guided-learning method suitable for networks of any size. The method employs a part of the weights for guiding and the other part for training and optimization. We demonstrate this technique by predicting residue solvent accessibility and real-value backbone torsion angles of proteins. In this application, the guiding factor is designed to satisfy the intuitive condition that for most residues, the contribution of a residue to the structural properties of another residue is smaller for greater separation in the protein-sequence distance between the two residues. We show that the guided-learning method makes a 2-4% reduction in 10-fold cross-validated mean absolute errors (MAE) for predicting residue solvent accessibility and backbone torsion angles, regardless of the size of database, the number of hidden layers and the size of input windows. This together with introduction of two-layer neural network with a bipolar activation function leads to a new method that has a MAE of 0.11 for residue solvent accessibility, 368 for w, and 228 for /.
dc.description.peerreviewedYes
dc.description.publicationstatusYes
dc.format.extent626083 bytes
dc.format.mimetypeapplication/pdf
dc.languageEnglish
dc.language.isoeng
dc.publisherWiley-Liss
dc.publisher.placeUnited States
dc.relation.ispartofstudentpublicationN
dc.relation.ispartofpagefrom847
dc.relation.ispartofpageto856
dc.relation.ispartofissue4
dc.relation.ispartofjournalProteins: Structure, Function, and Bioinformatics
dc.relation.ispartofvolume74
dc.rights.retentionY
dc.subject.fieldofresearchMathematical sciences
dc.subject.fieldofresearchBiological sciences
dc.subject.fieldofresearchInformation and computing sciences
dc.subject.fieldofresearchcode49
dc.subject.fieldofresearchcode31
dc.subject.fieldofresearchcode46
dc.titleImproving the prediction accuracy of residue solvent accessibility and real-value backbone torsion angles of proteins by guided-learning through a two-layer neural network
dc.typeJournal article
dc.type.descriptionC1 - Articles
dc.type.codeC - Journal Articles
gro.rights.copyright© 2009 Wiley Periodicals, Inc. This is the accepted version of the following article: Improving the prediction accuracy of residue solvent accessibility and real-value backbone torsion angles of proteins by guided-learning through a two-layer neural network, Proteins: Structure, Function, and Bioinformatics, Vol. 74(4), 2009, pp. 847-856, which has been published in final form at dx.doi.org/10.1002/prot.22193.
gro.date.issued2009
gro.hasfulltextFull Text
gro.griffith.authorZhou, Yaoqi


Files in this item

This item appears in the following Collection(s)

  • Journal articles
    Contains articles published by Griffith authors in scholarly journals.

Show simple item record