Naslov
Prepoznavanje pojačivača u genomu metodama strojnog učenja
(Application of machine learning methods to genome-wide prediction of enhancers)

Autori
Milisavljević, Alexandra

Vrsta, podvrsta i kategorija rada
Ocjenski radovi, diplomski rad, diplomski

Fakultet
Fakultet elektrotehnike i računarstva

Mjesto
Zagreb

Datum
30.06

Godina
2022

Stranica
56

Mentor
Domazet-Lošo, Mirjana

Ključne riječi
enhancers ; genome ; machine learning ; SVM ; MLP ; ensemble learning

Sažetak
Enhancers are cis-regulatory elements of DNA that positively regulate the transcription of their target genes. It is estimated that there are hundreds of thousands of enhancers in the human genome, which is too large a task to check experimentally. By applying machine learning methods, we could obtain models of adequate accuracy to predict if a DNA sequence is an enhancer or not. This thesis observes three classes of machine learning methods: support vector machine SVM, multilayer perceptron neural network MLP, and ensemble learning. The results of our models are then compared to existing tools for identifying enhancers, iEnhancer-2L and iEnhancer-EL. We also provide a visualization of the predicted results for the different models on the same DNA sequence.

Izvorni jezik
Engleski

Znanstvena područja
Računarstvo

POVEZANOST RADA