Mikrodizin kaynaklı genotip bilgisi kullanılarak kopya sayısı değişikliklerinde mozaiklik oranının saptanması

Abstract

Acun, H. M., Determination of Mosaicism Ratio in Copy Number Variations Using Microarray Based Genotype Information, Hacettepe University Graduate School of Health Sciences Department of Bioinformatics Master’s Thesis, Ankara, 2023. Copy number variations (CNVs) are >50 bp structural chromosomal variants that represent a regional change in the normal diploid (2 copies) copy number (CN) of genomic loci. This definition dictates that all copy numbers in a single cell must be integers, however, for a cell population consisting of various subpopulations with different copy numbers, when the whole cell population is considered there may be non-integer copy numbers. CNVs that cause such copy numbers to have been defined as mosaic CNVs (mCNVs). mCNVs can be encountered at birth or later in life because of conditions such as cancer. Since the detection of mCNVs is critical in diagnosis and treatment monitoring, especially in cancer, many researchers and commercial companies have developed methods for the detection of mCNVs. Among these methods, the SNP microarray method is the most commonly available as it allows genome-wide screening for mCNVs of different sizes. However, despite the rapid development in methods for localization of mCNVs, the number of competent and specific studies for determination of cells with CNV in a population of cells, fraction of mosaicism (fm), is very limited. Within this thesis, a bioinformatics tool which was named Copy-fm has been developed in R programming language that utilizes BAF and LRR values obtained from SNP microarray data to calculate ratio for fm by Kolmogorov-Smirnov (KS) test. This allowed us to fulfil the much-neglected need for fm calculation which may be required in research and clinics. Testing of Copy-fm's reliability and functionality has been achieved by using experimentally-designed mosaic deletion samples along with publicly available and real clinical data. In addition, Copy-fm has been tested on two main microarray platforms, revealing the differences between microarray platforms. In addition, besides the regional fm calculation, Copy-fm's ability to detect genome-wide mCNVs has been demonstrated by an approach we named “fm-first” algorithm, which has never been tested before. This thesis has been supported by TÜBİTAK-supported EJP-RD project RiboEurope (319S062) and Hacettepe Universty Research Projects Unit (THD-2021-19532).