Research Information System
Thesis Type: Doctorate
Institution Of The Thesis: Yildiz Technical University, Faculty Of Chemıcal And Metallurgıcal Engıneerıng, Department Of Bioengineering, Turkey
Approval Date: 2023
Thesis Language: English
Student: ÇAĞRI OĞUR
Supervisor: Alper Yılmaz
Abstract:
Balanced structural chromosome rearrangement (SR) carriers are poor prognosispatients in IVF due to the production of a high proportionof unbalanced gametesand a correspondingly low number of transferable embryos. Genetic and clinicaloutcomes are highly variable and noteasy to predict due to a high number of factors.There is an urgent need to identify these factors in order to offer these patients amore personalized treatment and improve their chances of having a baby. Since1963, a hypothesis entitled the “inter-chromosomal effect” (ICE) claimed that SRsinduce further chromosomal abnormalities among non-rearranged chromosomes ingametes/embryos. This hypothesis is mostly based on observation, lacks statisticalevidence and needs to be tested with analytical tools. The purpose of this studywas therefore to investigate factors that affect the proportion of chromosomallybalanced embryos available for transfer and test the “ICE” hypothesis. The endgoal was to establish predictive models using machine learning algorithms toimprove personalized treatments for SR carriers seeking preimplantation genetictesting (PGT-SR). Data analysis comprised genetic and clinical outcomes from 300couples referred to Şişli Memorial Hospital in vitro Fertilization (IVF) unit between2011-2019. A well-matched control group was selected from the same database.1835 embryos and 117,033 chromosome pairs were analyzed by array comperativegenomic hybridization (aCGH) and next generation sequencing (NGS). Statisticalanalysis was performed using SPSS and R-software. Rearrangement type, maternalage and sex of the carrier were found to havesignificant impacts on the proportionof transferable embryos. Results did not support any evidence for an ICE and thehypothesis was therefore rejected. This study helped to provide a predictive modelwith the use of advanced statistical and machine learning tools to reveal parametersto provide personalized treatment and better genetic counselling for PGT-SR couplesin the future.