Institutional Repository
Technical University of Crete
Technical University of Crete
EN
|
EL
| URI: | http://purl.tuc.gr/dl/dias/4D2829D5-6166-4220-9DFC-62EFA3A281D0 | ||
| Year | 2017 | ||
| Type of Item | Diploma Work | ||
| License |
|
||
| Bibliographic Citation | Dafni Andritsou, "Biclustering methods in gene analysis ", Diploma Work, Πολυτεχνείο Κρήτης: Σχολή Ηλεκτρολόγων Μηχανικών και Μηχανικών Υπολογιστών, Chania, Greece, 2017 https://doi.org/10.26233/heallink.tuc.68578 | ||
| Appears in Collections |
DNA microarrays comprise one of the most widespread experimental methods. This fact makes them a perfect tool for analysis and evaluation of cancer tissues, with the aim to find those molecular mechanisms governing the formation of tumors in different types of cancer and a further understanding of this disease. Molecular expression data are depicted in tables where the lines represent the genes and the columns represent the different experimental conditions.The aim of the different grouping techniques is to elicit crucial biological information regarding gene groups under specific circumstances. In this essay the method selected was the double categorization of gene expression data coming from samples of cell arrays of four different types of cancer (breast, ovarian, endometrial, cervical) gathered on a data table. The methodology proposed included the application of the Cheng & Church biclustering algorithm, by applying a series of revisions to it so as to elicit gene groups with similar patterns and a specific value range simultaneously. The number of genes (probes) included in the table rose to 33096, a fact that posed a serious burden to the study and analysis of the genes. For that reason, we brought the number of genes analyzed down to 1000 per type of cancer as well as focusing on studying the first three Biclusters each time. With this double categorization of data subspace we managed to create gene groups which showed genes with similar patterns along the cell arrays and, at the same time, their value range was small, making these groups more coherent and offering us the opportunity to elicit crucial biological information, such as the involvement of their genes in concrete biological processes (e.g., metabolism, macromolecular biosynthesis, reproduction) and specific molecular pathways (e.g., oxidative phosphorylation, RNA transfer, endocytosis).
Copyright © DIAS 2013
