Protein similarity networks and genetic algorithm driven feature selection for fold recognition

Valavanis, IK; Spyrou, GM; Nikita, KS

dc.contributor.author	Valavanis, IK	en
dc.contributor.author	Spyrou, GM	en
dc.contributor.author	Nikita, KS	en
dc.date.accessioned	2014-03-01T02:45:45Z
dc.date.available	2014-03-01T02:45:45Z
dc.date.issued	2008	en
dc.identifier.uri	https://dspace.lib.ntua.gr/xmlui/handle/123456789/32367
dc.subject	Cross Validation	en
dc.subject	Feature Selection	en
dc.subject	Fold Recognition	en
dc.subject	Genetic Algorithm	en
dc.subject	Machine Learning	en
dc.subject	Protein Sequence	en
dc.subject	Protein Structure	en
dc.subject	PSN	en
dc.subject	Random Walk	en
dc.subject.other	Cross validation	en
dc.subject.other	Feature selection	en
dc.subject.other	Fold recognition	en
dc.subject.other	Machine learning techniques	en
dc.subject.other	Protein sequences	en
dc.subject.other	Protein structures	en
dc.subject.other	Query proteins	en
dc.subject.other	Random Walk	en
dc.subject.other	Similarity network	en
dc.subject.other	Testing sets	en
dc.subject.other	Bioinformatics	en
dc.subject.other	Genetic algorithms	en
dc.subject.other	Learning algorithms	en
dc.subject.other	Optimization	en
dc.subject.other	Statistical tests	en
dc.subject.other	Set theory	en
dc.title	Protein similarity networks and genetic algorithm driven feature selection for fold recognition	en
heal.type	conferenceItem	en
heal.identifier.primary	10.1109/BIBE.2008.4696704	en
heal.identifier.secondary	http://dx.doi.org/10.1109/BIBE.2008.4696704	en
heal.identifier.secondary	4696704	en
heal.publicationDate	2008	en
heal.abstract	Fold recognition based on sequence-derived features is a complex classification problem and usuall sequence-derived features are exploited using proper machine learning techniques. Here we adress the task of fold recognition on a protein similarity network (PSN) basis. We construct a protein sequence similarity network (PSeSN) using a set of 125 sequence-derived features for an available set of 311 proteins.PSeSN is optimized by using a Genetic Algorithm (GA) to select the features that construct a PSeSN which is as similar as possible with the corresponding protein structure similarity network (PStSN). A random walk based algorithm is then utilized to recognize the fold of a query protein sequence by calculating its affinities to sequences-vertices both in the initial and the optimized PSeSN. Total accuracy (TA) measurements obtained using 10-fold cross validation show that the use of 48 out of 125 sequence-derived features (optimized PSeSN) yielded better results (mean TA: 0.35 in testing sets) than the initial PSeSN (mean TA: 0.316 in testing sets).	en
heal.journalName	8th IEEE International Conference on BioInformatics and BioEngineering, BIBE 2008	en
dc.identifier.doi	10.1109/BIBE.2008.4696704	en