Open Access Open Badges Research article

Ensemble approach combining multiple methods improves human transcription start site prediction

David G Dineen12*, Markus Schröder3, Desmond G Higgins2 and Pádraig Cunningham1

Author Affiliations

1 Complex and Adaptive Systems Laboratory (CASL), University College Dublin, Belfield, Dublin 4, Ireland

2 The Conway Institute of Biomolecular and Biomedical Research, University College Dublin, Belfield, Dublin 4, Ireland

3 BRF, Center for Biotechnology (CeBiTec), Bielefeld University, Bielefeld, Germany

For all author emails, please log on.

BMC Genomics 2010, 11:677  doi:10.1186/1471-2164-11-677

Published: 30 November 2010



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.