prikaz prve stranice dokumenta Predviđanje biološke aktivnosti malih molekula modelom strojnog učenja
  Download
PDF 1.99 MB
undergraduate thesis
Predviđanje biološke aktivnosti malih molekula modelom strojnog učenja
2023. urn:nbn:hr:190:484829
Tomasić, Jakov
University of Rijeka
Faculty of Engineering
Department of Computer Engineering
Section of Intelligent Computing Systems

Cite this document

Tomasić, J. (2023). Predviđanje biološke aktivnosti malih molekula modelom strojnog učenja (Undergraduate thesis). Rijeka: University of Rijeka, Faculty of Engineering. Retrieved from https://urn.nsk.hr/urn:nbn:hr:190:484829

Tomasić, Jakov. "Predviđanje biološke aktivnosti malih molekula modelom strojnog učenja." Undergraduate thesis, University of Rijeka, Faculty of Engineering, 2023. https://urn.nsk.hr/urn:nbn:hr:190:484829

Tomasić, Jakov. "Predviđanje biološke aktivnosti malih molekula modelom strojnog učenja." Undergraduate thesis, University of Rijeka, Faculty of Engineering, 2023. https://urn.nsk.hr/urn:nbn:hr:190:484829

Tomasić, J. (2023). 'Predviđanje biološke aktivnosti malih molekula modelom strojnog učenja', Undergraduate thesis, University of Rijeka, Faculty of Engineering, accessed 31 March 2025, https://urn.nsk.hr/urn:nbn:hr:190:484829

Tomasić J. Predviđanje biološke aktivnosti malih molekula modelom strojnog učenja [Undergraduate thesis]. Rijeka: University of Rijeka, Faculty of Engineering; 2023 [cited 2025 March 31] Available at: https://urn.nsk.hr/urn:nbn:hr:190:484829

J. Tomasić, "Predviđanje biološke aktivnosti malih molekula modelom strojnog učenja", Undergraduate thesis, University of Rijeka, Faculty of Engineering, Rijeka, 2023. Available at: https://urn.nsk.hr/urn:nbn:hr:190:484829

Please login to the repository to save this object to your list.
Metadata
TitlePredviđanje biološke aktivnosti malih molekula modelom strojnog učenja
Title (english)Machine learning-based biological activity prediction for small molecules
AuthorJakov Tomasić
MentorGoran Mauša (mentor)
Committee memberMiroslav Joler (predsjednik povjerenstva)
Committee memberGoran Mauša (član povjerenstva)
GranterUniversity of Rijeka
Faculty of Engineering
(Department of Computer Engineering)
(Section of Intelligent Computing Systems)
Rijeka
Defense date and country2023-09-12, Croatia
Scientific / art field,
discipline and subdiscipline		TECHNICAL SCIENCES
Computing
Abstract
Bitan dio moderne medicine je predviđanje reakcija raznih molekula i mikroorganizama. Ovaj rad istražuje rješenje predviđanja reakcije malih molekula i mikroorganizama iz baze otvorenog pristupa CO-ADD. Korištene su razne tehnike za poboljšanje rezultata: skaliranje i centriranje ulaznih podataka, brisanje više zastupljenih podataka, dupliciranje rijetkih slučajeva, pretvorba problema označavanja u problem klasifikacije, algoritamska pretraga najboljih hiperparametara modela, uklanjanje... More kontradiktornih podataka te treniranje više puta. Čak i nakon primjene svih ovih tehnika dobiveni rezultati su daleko od optimalnih. Korištena baza podataka u kombinaciji s knjižnicom Mordred za pretvaranje SMILES zapisa molekula u njihove značajke nema dovoljno podataka za kvalitetno rješenje ovog problema. Uočena je velika prednost u performansama tehnike klasifikacije s višestrukim oznakama u odnosu na češće korištenu tehniku binarne klasifikacije. Ta prednost nosi i nedostatke poput teže obrade podataka i malo lošijih rezultata. Više istraživanja potrebno je za kvalitetniju usporedbu ove dvije tehnike. Najkorisnija tehnika za predobradu podataka bila je principal component analysis (PCA) koja je smanjila broj ulaznih podataka i time omogućila rezultate sa srednjom vrijednosti F1 metrike jednakom 0.429 za klasifikaciju s višestrukim oznakama, odnosno 0.438 za binarnu klasifikaciju. Ovaj rad utvrđuje korisnost tehnike višestruke klasifikacije s ciljem manje potrošnje resursa, ali i pokazuje težinu problema predviđanja biološke reakcije malih molekula i mikroorganizama. Less
Abstract (english)
One of the major parts of modern medicine research is detecting reactions of various molecules and microorganisms. This paper explores a solution for predicting the reactions of small molecules and microorganisms from the open-access CO-ADD database. Various techniques for improving the results were used: scaling and centering the input data, removing overrepresented data, duplicating rare cases, transforming the labeling problem into a classification problem, algorithmic search for... More the best model hyperparameters, removing contradictory data, and training multiple times. Even after applying all of these techniques, the obtained results are far from optimal. The data from the used database and the Mordred toolkit for converting SMILES representations of molecules into their features isn’t sufficient for a high-quality solution to this problem. This paper shows a significant advantage in the performance of the multi-label classification technique compared to the more conventional binary classification technique. This advantage is followed by its drawbacks like difficulties with data processing and slightly poorer results. Further research is needed for a more robust comparison of these two techniques. The most useful data preprocessing technique was principal component analysis (PCA), which reduced the number of data inputs and thus enabled results with a median F1 score of 0.429 for multi-label classification, and 0.438 for binary classification. This paper confirms the usefulness of multi-task learning with the goal of lower resource usage while also demonstrating the complexity of predicting the biological reactions of small molecules and microorganisms Less
Keywords
Keywords (english)
Languagecroatian
URN:NBNurn:nbn:hr:190:484829
Study programmeTitle: Computer Engineering
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 resourceText
File originBorn digital
Access conditionsOpen access
Terms of useLicence Attribution 4.0 International (CC BY 4.0) icon
Created on2023-10-24 11:15:36
accessibility

closeAccessibilityrefresh

If you wish to permanently save changes, click on Save, if not - your settings will be reset when you restart the browser.