Distance matrices, reduced hidden Markov models and protein secondary structure comparison

Zagorščak, Maja (2012) Distance matrices, reduced hidden Markov models and protein secondary structure comparison. Diploma thesis, Faculty of Science > Department of Biology.

[img] PDF
Restricted to Registered users only
Language: Croatian

Download (4MB) | Request a copy

Abstract

This work discusses recognition of protein supra-secondary structure by combining two methods: distance matrices (DM) and reduced profle Hidden Markov Models (rHMMp). Distance matrix method was frst used to detect supra-secondary structure elements in polypeptide chains. This was used to build a model (rHMMp), and an optimal model- sequence assignment was generated using Position Specifc Scoring Matrix (PSSM) and inverse posterior assignment (IPA) techniques. Results show a clear diference in optimal scores between sequences with supra-secondary structure elements and those without. It can be seen that, on data-sets used, structure similarity has been successfully recognized as distant homology at primary structure level.

Item Type: Thesis (Diploma thesis)
Keywords: primary and supra-secondary protein structure, motifs, protein distance matrix, distance matrix for secondary structure pairs; inverse posterior assignment, PSSM
Supervisor: Goldstein, Pavle
Date: 2012
Number of Pages: 74
Subjects: NATURAL SCIENCES > Biology
Divisions: Faculty of Science > Department of Biology
Depositing User: Grozdana Sirotic
Date Deposited: 29 May 2014 11:20
Last Modified: 22 Sep 2014 08:25
URI: http://digre.pmf.unizg.hr/id/eprint/1922

Actions (login required)

View Item View Item

Nema podataka za dohvacanje citata