Preciznost i klasifikacija

master's thesis

• Author: Monika Mirković
• Mentor: Pavle Goldstein (mentor)
• Committee member: Pavle Goldstein (predsjednik povjerenstva)
• Committee member: Siniša Slijepčević (član povjerenstva)
• Committee member: Dijana Ilišević (član povjerenstva)
• Committee member: Miljenko Huzak (član povjerenstva)
• Granter: University of Zagreb; Faculty of Science; (Department of Mathematics); Zagreb
• Defense date and country: 2020-02-27, Croatia
• Scientific / art field, discipline and subdiscipline: NATURAL SCIENCES; Mathematics

Abstract

Razvojem bioinformatike i otkrivanjem novih proteina javlja se potreba za razvojem novih metoda klasifikacije proteina. U ovom diplomskom radu promatrali smo kraće nizove aminokiselina s karakterističnim mutacijama koje nazivamo motivi. Opisali smo postupak traženja karakterističnih motiva za neku proteinsku familiju. Posebno je važno istaknuti da traženi motivi nisu fiksne duljine. Zatim smo opisali način ocjenjivanja proteina u odnosu na profile dobivenih karakterističnih motiva pomoću kojeg provodimo klasifikaciju proteina u proteinske familije. Kako bi provjerili točnost našeg algoritma proveli smo postupak traženja karakterističnih motiva na dvije proteinske familije, lipoproteinskim lipazama i Walkerovim motivima. Klasifikacijom proteina iz test setova obje familije dobiveni su iznenađujuće dobri rezultati.

Abstract (english)

With the advancement of bioinformatics and the discovery of new proteins, a need for new protein classification methods has arisen. This master’s thesis focuses on the short sequences of amino acids with characteristic mutations called motifs. We describe a process of characteristic motif discovery for a given protein family. It is worth mentioning that the observed motifs are of variable length. Further, we describe the evaluation of proteins with respect to the profiles of the discovered characteristic motifs. Based on this, we classify the proteins in protein families. In order to validate the algorithm, we apply it to two protein families: lipoprotein lipase family and Walker motifs. The results of the protein classification for the test sets from both families are surprisingly good.

• Keywords: bioinformatika; klasifikacija proteina; motivi
• Keywords (english): bioinformatics; protein classification; motifs
• Language: croatian
• URN:NBN: urn:nbn:hr:217:297065
• Study programme: Title: Mathematical Statistics; Study programme type: university; Study level: graduate; Academic / professional title: magistar/magistra matematike (magistar/magistra matematike)
• Type of resource: Text
• File origin: Born digital
• Access conditions: Open access
• Created on: 2020-09-21 10:34:02