University Links: Home Page | Site Map
Covenant University Repository

A MACHINE LEARNING MODEL FOR PREDICTING ESSENTIAL GENES FROM PLASMODIUM FALCIPARUM METABOLIC NETWORK

STEPHEN, NPOANDAN BINAANSIM and Covenant University, Theses (2023) A MACHINE LEARNING MODEL FOR PREDICTING ESSENTIAL GENES FROM PLASMODIUM FALCIPARUM METABOLIC NETWORK. Masters thesis, COVENANT UNIVERSITY.

[img] PDF
Download (1MB)

Abstract

The accurate prediction of essential metabolic genes (i.e., genes necessary for cell survival) in eukaryotic organisms is still a difficult task in bioinformatics, especially in pathogenic species like Plasmodium falciparum, the malaria causing parasite. The difficulty and cost in time and resources of experimental methods has necessitated the use of computational methods for gene essentiality prediction. The majority of earlier research in this field concentrated on prokaryotes, omitting the complexity of weighted and directed metabolite transport in metabolic networks. To overcome this limitation, we developed a Network-based Machine Learning framework that explored various network properties in Plasmodium falciparum using the Genome-Scale Metabolic Model (iAM_Pf480) adopted from the BiGG database and essentiality data from the Ogee database. Our machine learning framework significantly increased the accuracy of gene essentiality predictions by taking into account the weighted and directed nature of the metabolic network and utilising network-based features, producing state-of-the-art results with an accuracy of 0.85 and AuROC of 0.7. This study expanded our knowledge of the complex nature of metabolic networks and their critical function in determining the essentiality of genes. Notably, our model identified important genes that were previously classified as non-essential in the Ogee database but predicted to be essential. Some of these genes have previously been linked to potential drug targets for the treatment of malaria, providing promising new research directions.

Item Type: Thesis (Masters)
Uncontrolled Keywords: Gene essentiality, Constraint-based Analysis, Graph-based analysis, Machine Learning, Metabolic Networks
Subjects: Q Science > Q Science (General)
Q Science > QA Mathematics > QA75 Electronic computers. Computer science
Divisions: Faculty of Engineering, Science and Mathematics > School of Electronics and Computer Science
Depositing User: AKINWUMI
Date Deposited: 03 Oct 2023 13:31
Last Modified: 03 Oct 2023 13:31
URI: http://eprints.covenantuniversity.edu.ng/id/eprint/17337

Actions (login required)

View Item View Item