PhD project

Christoffer Trier Månsson, MSs

Publications (link to AU profile)

I have a bachelor’s and a master’s degree in Molecular Medicine from Aarhus University. As of January 2022, I have been enrolled as PhD student at the Graduate School of Health at Aarhus University. My main project is to develop and use cell-free chromatin immunoprecipitation sequencing (cfChIP-seq) on plasma samples from lung cancer patients. We use cfChIP-seq to estimate tumor gene expression based on liquid biopsies.

cfChIP utilizes that circulating cell-free DNA (cfDNA) is bound to nucleosomes which contain post-translational modifications representative of the epigenetics in the cell of origin. The epigenetic marker, histone 3 lysine-36 trimethylation (H3K36me3), is associated with active transcription. Immunoprecipitation of H3K36me3 modified nucleosomes from plasma followed by next-generation sequencing (NGS) or droplet digital PCR (ddPCR) can be used to quantify gene activity in lung cancer patients. We call this approach cfChIP, and we have used it to study gene expression profiles in lung cancer patients of different histological and molecular subtypes.

In addition, I am very interested in cfDNA fragmentomics and how this field can be used to expand the utility of liquid biopsies. Recently, we have published a paper describing how different cfDNA fragmentomic features, including fragment length and fragment end motifs, are associated with gene expression. I also believe that cfDNA fragmentomics can be used to increase the sensitivity of ctDNA detection. We have recently published a paper where we use in vitro size selection to increase the mutational allele fraction of ctDNA and increase the number of copy-number alterations detected in plasma.

Besides this, I also work with variant call after NGS which has led to the development of the R package DNAfusion. DNAfusion takes a BAM file as input and can differentiate between EML4-ALK positive and negative samples. It has been developed to work downstream of the AVENIO pipeline but any BAM format generated using a hybridization-capture NGS can work as input. In our publication from 2023 we show how DNAfusion identifies more EML4-ALK positive samples than AVENIO alone.