Predviđanje napredovanja infekcije HIV-om na temelju sljedova virusne RNA

undergraduate thesis

• Title (english): Predicting the progression of HIV infection based on viral RNA sequences
• Author: Toni Ivanković
• Mentor: Mirjana Domazet-Lošo (mentor)
• Committee member: Mirjana Domazet-Lošo (predsjednik povjerenstva)
• Committee member: Ivica Botički (član povjerenstva)
• Committee member: Tomislav Jagušt (član povjerenstva)
• Granter: University of Zagreb; Faculty of Electrical Engineering and Computing; Zagreb
• Defense date and country: 2022-07-05, Croatia
• Scientific / art field, discipline and subdiscipline: TECHNICAL SCIENCES; Computing

Abstract

Zbog relativno velike vremenske složenosti algoritma poravnanja sljedova RNA, pojavljuje se potreba za alternativnim analizama koje se obavljaju u manjoj vremenskoj složenosti. Analiza k-mera iz slijeda RNA izdvaja moguće podnizove nukleotida te promatra uzorke u njihovom broju pojavljivanja. Za predviđanje traženog parametra korištena je duboka potpuno povezana umjetna neuronska mreža. Izazov u radu predstavljao je nebalansiran skup podataka za treniranje, u kojemu je instanci jedne klase bilo značajno više od onih druge klase. Najbolji rezultati dobiveni su za duljinu k-mera jednaku 5, ali čak i najbolji rezultati, u usporedbi s ostalim postojećim metodama, pokazuju da je potrebno žrtvovati kvalitetu predviđanja radi smanjenja vremenske složenosti.

Abstract (english)

Because of the relatively high time complexity of the RNA sequence alignment algorithm, there comes a need for alternate analyses which operate in a lower time complexity. K-mer analysis extracts possible nucleotide subsequences from an RNA sequence and observes patterns in their number of appearances. To predict the sought parameter, a deep fully connected artificial neural network was used. A challenge was imposed by the imbalanced training dataset, which contained significantly more instances of one class, compared to the other class. The best results were achieved for k-mer size equal to 5, but in comparison to other existing methods, even the best achieved results show that it is necessary to sacrifice the prediction quality in order to reduce the time complexity.

• Keywords: strojno učenje; umjetne neuronske mreže; HIV; k-mer; F1-mjera; nebalansiran skup podataka; TensorFlow; prenaučenost; analiza osjetljivosti
• Keywords (english): machine learning; artificial neural networks; HIV; k-mer; F1-metric; imbalanced dataset; TensorFlow; overfitting; sensitivity analysis
• Language: croatian
• URN:NBN: urn:nbn:hr:168:462006
• Study programme: Title: Computing; Study programme type: university; Study level: undergraduate; Academic / professional title: sveučilišni/a prvostupnik/ prvostupnica (baccalaureus/baccalaurea) inženjer/inženjerka računarstva (sveučilišni/a prvostupnik/ prvostupnica (baccalaureus/baccalaurea) inženjer/inženjerka računarstva)
• Type of resource: Text
• File origin: Born digital
• Access conditions: Closed access
• Created on: 2023-01-15 10:21:54