Uporaba metoda nenadziranog učenja za potrebe kemoinformatičkih analiza

master's thesis

• Title (english): The use of unsupervised learning methods for cheminformatics analyses
• Author: Marko Kremer
• Mentor: Damir Pintar (mentor)
• Committee member: Damir Pintar (predsjednik povjerenstva)
• Committee member: Mihaela Vranić (član povjerenstva)
• Committee member: Luka Humski (član povjerenstva)
• Granter: University of Zagreb; Faculty of Electrical Engineering and Computing; Zagreb
• Defense date and country: 2024-07-01, Croatia

Abstract

Kemoinformatika je interdisciplinarno područje koje spaja kemiju sa računalnom odnosno informacijskom znanošću, koristeći računarske tehnike za rješavanje kemijskih problema. Jedan od elemenata kemoinformatike su molekularni otisci. Molekularni otisci su binarni zapisi strukture molekula. Molekularni otisci korišteni u ovom radu su MACCS, Morgan, Atom Pair, Pharmacore, Topological Torsions i RDKit otisak. Za uspoređivanje molekula koristilo se nekoliko metoda sličnosti: Tanimotov koeficijent, Euklidska udaljenost, Manhattanova udaljenost, kosinusna sličnost, Dice koeficijent i Russel-RAO koeficijent. Nakon oddabira otiska i metode sličnosti provelo se grupiranje Taylor-Butina algoritmom, koji je napravljen posebno za područje kemoinformatike.

Abstract (english)

Chemoinformatics is an interdisciplinary field that combines chemistry with computer or information science, using computer techniques to solve chemical problems. One of the elements of chemoinformatics is molecular fingerprints. Molecular fingerprints are binary records of the structure of molecules. Molecular fingerprints used in this work are MACCS, Morgan, Atom Pair, Pharmacore, Topological Torsions and RDKit fingerprint. Several similarity methods were used to compare molecules: Tanimoto coefficient, Euclidean distance, Manhattan distance, cosine similarity, Dice coefficient and Russel-RAO coefficient. After selecting the fingerprint and the similarity method, clustering was carried out using the Taylor-Butin algorithm, which was created specifically for the field of chemoinformatics.

• Keywords: kemoinformatika; molekularni otisci; nenadzirano učenje; grupiranje; strojno učenje; mjere sličnosti
• Keywords (english): chemoinformatics; molecular fingerprints; unsupervised learning; clustering; machine learning; similarity measures
• Language: croatian
• URN:NBN: urn:nbn:hr:168:547239
• Study programme: Title: Computing; Study programme type: university; Study level: graduate; Academic / professional title: magistar/magistra inženjer/inženjerka računarstva (magistar/magistra inženjer/inženjerka računarstva)
• Type of resource: Text
• File origin: Born digital
• Access conditions: Closed access
• Created on: 2024-10-12 21:11:47