Ensemble approach combining multiple methods improves human transcription start site prediction.

Hdl Handle:
http://hdl.handle.net/10147/125892
Title:
Ensemble approach combining multiple methods improves human transcription start site prediction.
Authors:
Dineen, David G; Schröder, Markus; Higgins, Desmond G; Cunningham, Pádraig
Affiliation:
Complex and Adaptive Systems Laboratory (CASL), University College Dublin, Belfield, Dublin 4, Ireland. david.dineen@ucd.ie
Citation:
Ensemble approach combining multiple methods improves human transcription start site prediction. 2010, 11:677 BMC Genomics
Journal:
BMC genomics
Issue Date:
2010
URI:
http://hdl.handle.net/10147/125892
DOI:
10.1186/1471-2164-11-677
PubMed ID:
21118509
Abstract:
The computational prediction of transcription start sites is an important unsolved problem. Some recent progress has been made, but many promoters, particularly those not associated with CpG islands, are still difficult to locate using current methods. These methods use different features and training sets, along with a variety of machine learning techniques and result in different prediction sets.; We demonstrate the heterogeneity of current prediction sets, and take advantage of this heterogeneity to construct a two-level classifier ('Profisi Ensemble') using predictions from 7 programs, along with 2 other data sources. Support vector machines using 'full' and 'reduced' data sets are combined in an either/or approach. We achieve a 14% increase in performance over the current state-of-the-art, as benchmarked by a third-party tool.; Supervised learning methods are a useful way to combine predictions from diverse sources.
Item Type:
Article
Language:
en
MeSH:
Base Pairing; Computational Biology; Genome, Human; Humans; Principal Component Analysis; Promoter Regions, Genetic; Software; Transcription Initiation Site
ISSN:
1471-2164

Full metadata record

DC FieldValue Language
dc.contributor.authorDineen, David Gen
dc.contributor.authorSchröder, Markusen
dc.contributor.authorHiggins, Desmond Gen
dc.contributor.authorCunningham, Pádraigen
dc.date.accessioned2011-03-28T14:50:53Z-
dc.date.available2011-03-28T14:50:53Z-
dc.date.issued2010-
dc.identifier.citationEnsemble approach combining multiple methods improves human transcription start site prediction. 2010, 11:677 BMC Genomicsen
dc.identifier.issn1471-2164-
dc.identifier.pmid21118509-
dc.identifier.doi10.1186/1471-2164-11-677-
dc.identifier.urihttp://hdl.handle.net/10147/125892-
dc.description.abstractThe computational prediction of transcription start sites is an important unsolved problem. Some recent progress has been made, but many promoters, particularly those not associated with CpG islands, are still difficult to locate using current methods. These methods use different features and training sets, along with a variety of machine learning techniques and result in different prediction sets.-
dc.description.abstractWe demonstrate the heterogeneity of current prediction sets, and take advantage of this heterogeneity to construct a two-level classifier ('Profisi Ensemble') using predictions from 7 programs, along with 2 other data sources. Support vector machines using 'full' and 'reduced' data sets are combined in an either/or approach. We achieve a 14% increase in performance over the current state-of-the-art, as benchmarked by a third-party tool.-
dc.description.abstractSupervised learning methods are a useful way to combine predictions from diverse sources.-
dc.language.isoenen
dc.subject.meshBase Pairing-
dc.subject.meshComputational Biology-
dc.subject.meshGenome, Human-
dc.subject.meshHumans-
dc.subject.meshPrincipal Component Analysis-
dc.subject.meshPromoter Regions, Genetic-
dc.subject.meshSoftware-
dc.subject.meshTranscription Initiation Site-
dc.titleEnsemble approach combining multiple methods improves human transcription start site prediction.en
dc.typeArticleen
dc.contributor.departmentComplex and Adaptive Systems Laboratory (CASL), University College Dublin, Belfield, Dublin 4, Ireland. david.dineen@ucd.ieen
dc.identifier.journalBMC genomicsen

Related articles on PubMed

All Items in Lenus, The Irish Health Repository are protected by copyright, with all rights reserved, unless otherwise indicated.