Doctoral defence: Priit Paluoja "Computational methods for NIPT-based aneuploidy and microdeletion screening"

On 25 August at 10:00 Priit Paluoja will defend his thesis "Computational methods for NIPT-based aneuploidy and microdeletion screening".

Supervisors:
Professor Andres Salumets, University of Tartu
Professor Priit Palta, Estonian Genome Centre
juhatuse liige Kaarel Krjutškov, Celvia CC AS

Opponent:
Silvia Bottini, National Research Institute for Agriculture, Food and the Environment (France)

Summary
Non-invasive prenatal genetic testing (NIPT) is a screening method for assessing the risk of fetal chromosomal disorders from a maternal blood sample. NIPT is based on sequencing and data analysis of cell-free DNA originating from the placenta. In addition to detecting changes in the fetal chromosome copy number, NIPT can also identify the risk of pathogenic microdeletions. A microdeletion syndrome results from the loss of a small segment of a chromosome, and its clinical severity depends on the deleted region. For example, a microdeletion in the 22q11 region causes DiGeorge syndrome, which is associated with heart defects, cleft palate, and intellectual disabilities. Fetal chromosomal disorder risk assessment is a bioinformatics challenge, as the accurate analysis of millions of DNA sequences requires sophisticated computational solutions.

The aim of this doctoral dissertation is to improve the accuracy, data utilization, and cost-effectiveness of NIPT prenatal screening. A new computational framework for NIPT was developed, and previously published NIPT algorithms were validated and analyzed using clinical data. Additionally, a novel software tool, BinDel, was developed and clinically validated for estimating fetal microdeletion risk.

The results quantified the accuracy of computational tools under varying sequencing coverage and fetal DNA fraction levels. The findings showed that NIPT accuracy is influenced by both sequencing depth and the fetal DNA fraction, as well as the choice of algorithm. The new BinDel software improved the ability to detect microdeletions, providing potential for more accurate and widespread prenatal screening. Additionally, a machine learning-based computational framework was developed for detecting fetal chromosomal copy number changes.

You can also watch defense via Teams.