Promoter Prediction Using Physico-Chemical Properties of DNA

Abstract

Abstract. The ability to locate promoters within a section of DNA is known to be a very difficult and very important task in DNA analysis. We document an approach that incorporates the concept of DNA as a complex molecule using several models of its physico-chemical properties. A support vector machine is trained to recognise promoters by their distinctive physical and chemical properties. We demonstrate that by combining models, we can improve upon the classification accuracy obtained with a single model. We also show that by examining how the predictive accuracy of these properties varies over the promoter, we can reduce the number of attributes needed. Finally, we apply this method to a real-world problem. The results demonstrate that such an approach has significant merit in its own right. Furthermore, they suggest better results from a planned combined approach to promoter prediction using both physicochemical and sequence based techniques.

Machine learning based prediction for peptide drift times in ion mobility spectrometry

“…The physicochemical properties of a peptide/protein sequence have been successfully used to predict subcellular localization of protein sequences (Garg et al, 2009;Tantoso and Li, 2008), protein-protein interactions (Agrawal et al, 2005;Bock and Gough, 2001;Nanni and Lumini, 2006), promoter regions (Uren et al, 2006), long disordered regions (Hirose et al, 2007) and sequence homology (Yang et al, 2008) among others. Our selection of these properties is a variant of a model used for predicting proteotypic peptides (peptides that are likely to be observed in MS-based experiments) (Webb-Robertson et al, 2008).…”

Generic eukaryotic core promoter prediction using structural features of DNA

“…It has indeed been shown that sequence properties such as GC content and more general chemo-physical properties of the DNA, such as stabilizing energy of Z-DNA (Ho et al 1990), DNA denaturation values (Blake and Delcourt 1998;Blake et al 1999), protein-induced deformability (Olson et al 1998), and duplex-free energy (Sugimoto et al 1996), among others (for review, see Florquin et al 2005), can be used to describe (core) promoters, and to discriminate between (core) promoter sequences and non (core) promoter sequences Florquin et al 2005;Kanhere and Bansal 2005;Uren et al 2006;Wang and Benham 2006). Because all these properties are calculated from conversion tables using di-or trinucleotides, one may argue that these properties are in fact exactly the same as the nucleotide sequence and do not offer any additional information.…”

Inferring Phylogenies from Physico-Chemical Properties of DNA

“…Utilization of reduced alphabets based on physicochemical properties of DNA or amino acid sequences have been assessed by bioinformaticians and molecular biologists for analysing binding domains [14], promoter prediction [15,16], gene finding [17], Nonsynonymous Mutations in the Protein-Coding Genes [18] and many others. However, effect of physico-chemical properties on phylogeny construction was not very well covered within these studies.…”