On Tuesday 22nd April 2025, Dat Nguyen successfully defended his PhD thesis, entitled “Improved computational methods for regulatory genomic inference”.
The overarching aim of the PhD project was to create more accurate, efficient and scalable tools to link genetic variations to regulation of gene expression, especially in non-coding regions of the genome.
Although there are millions of genetic differences between individuals in humans and other animal species, only a few such differences have real significance when it comes to affecting biology. Dat’s PhD has concentrated on parts of the DNA that do not code for proteins but may still have the ability to regulate how genes function – an area that is difficult to interpret. A major result has been the development of new computational tools that help researchers identify the genetic changes that affect how an organism functions at the molecular level. One tool improves the analysis of circular RNA, a type of RNA that plays an important role in gene regulation, making it easier to detect genetic variations that influence this RNA type. Another tool enables the detection of links between genetic variants and molecular data even with data from a small number of individuals, which helps reduce the cost of such experiments. A third tool uses artificial intelligence to predict how DNA changes might affect gene regulation in livestock and fish.
Dat presenting his trial lecture on deep learning methods for computational biology
Dat’s defence also featured a trial lecture on the subject “State of the art in Deep Learning methods for computational biology”.
Dat’s PhD was supervised by CIGENE’s Lars Grønvold, and co-supervised by Sigbjørn Lien and Simen Rød Sandve (both CIGENE). The examination committee comprised Associated Professor Kaur Alasoo (University of Tartu, Estonia), Professor Serap Gonen (Benchmark genetics), and internal examiner Dr. Gareth Difford (NMBU).
The doctoral thesis is available for public review on this link (download): https://filesender.sikt.no/?s=download&token=bd03ed56-39e6-4315-b871-cd0cb4139c46