master's thesis
Selma Halilčević (Author), Aleksander Sadikov (Mentor), Barbara Breznik (Co-mentor)

Abstract

Brain tumour glioblastoma (GBM) is one of the most aggressive, invasive, and unfortunately lethal tumours. The unfavourable prognosis of GBM has encouraged continued efforts to better understand its pathobiology to find more clinically relevant biomarkers and novel efficient therapeutic approaches. Treatment of GBM remains one of the hardest challenges in cancer therapy, firstly due to the resistance to therapy of glioblastoma stem cells and secondly due to the tumour heterogeneity, leading to variable treatment responses. Computer scientists join their forces with biological scientists and clinicians to bring GBM research to a next level. In this master thesis, we perform an analysis of glioblastoma stem cells and other GBM-related biomarker gene expressions. We analyse correlations of gene expression of several selected markers with clinical data, and their significance for survival of GBM patients. We find several biomarkers that can be used as prognostic biomarkers. With the use of machine learning methods, we define a new approach to determine GBM subtypes. We prove the existence of a new MIX subtype, which contains all GBM subtypes (classical, mesenchymal and proneural) due to the high GBM heterogeneity.

Keywords

glioblastoma;biomarkers;cancer stem cells;correlations;classification;survival analysis;machine learning;computer science;master's thesis;

Data

Language: English
Year of publishing:
Typology: 2.09 - Master's Thesis
Organization: UL FRI - Faculty of Computer and Information Science
Publisher: [S. Halilčević]
UDC: 004.85:616-006(043.2)
COBISS: 177581315 Link will open in a new window
Views: 36
Downloads: 5
Average score: 0 (0 votes)
Metadata: JSON JSON-RDF JSON-LD TURTLE N-TRIPLES XML RDFA MICRODATA DC-XML DC-RDF RDF

Other data

Secondary language: Slovenian
Secondary title: Analiza označevalcev rakavih matičnih celic kot prognostičnih označevalcev pri glioblastomu
Secondary abstract: Možganski tumor, glioblastom, je eden izmed najbolj agresivnih, invazivnih in žal smrtonosnih tumorjev. Slaba prognoza glioblastoma je spodbudila raziskave, da bi bolj razumeli patobiologijo glioblastoma in našli klinično biološke označevalce in nove načine zdravljenja. Zdravljenje glioblastoma je eden od največjih izzivov v onkologiji zaradi odpornosti glioblastomskih matičnih celic na terapijo ter heterogenosti glioblastoma. V tej študiji, smo naredili analizo izražanja genov bioloških označevalcev rakavih matičnih celic in ostalih označevalcev, ki so povezani z napredovanjem glioblastoma. Naredili smo analizo korelacij med genskim izražanjem več bioloških označevalcev in kliničnih podatkov, ter analizo njihove vloge pri napovedanju preživetja bolnikov z glioblastomom. Ugotovili smo, da nekateri biološki označevalci se lahko uporabljajo kot prognostični biološki označevalci. Z uporabo metod strojnega učenja smo definirali nov pristop za določitev podtipov glioblastoma. Dokazali smo prisotnost novega podtipa glioblastoma (podtip MIX), ki združuje vse ostale podtipe (klasični, mezenhimski in proneuralni) zaradi velike heterogenosti glioblastoma.
Secondary keywords: glioblastom;biološki označevalci;rakave matične celice;korelacije;klasifikacija;analiza preživetja;magisteriji;Rak (bolezen);Onkologija;Strojno učenje;Računalništvo;Univerzitetna in visokošolska dela;
Type (COBISS): Master's thesis/paper
Study programme: 1000471
Embargo end date (OpenAIRE): 1970-01-01
Thesis comment: Univ. v Ljubljani, Fak. za računalništvo in informatiko
Pages: XX, 117 str.
ID: 21492990