About me
Felix Hofmann
Master-student
fehofman@students.uni-mainz.de
Current projects
Machine learning based Pseudouridine detection from Oxford Nanopore sequencing data
RNA modifications are known to be essential for major biological processes including immune function, cell differentiation and stress response. Thus, they are increasingly associated to an ever-growing number of diseases. According to the MODOMICS database, there are about 200 known RNA modifications associated with 175 different diseases in humans. Among this list, one of the most abundant RNA modifications is pseudouridine, which is present at ~0.2–0.6% of total U in mRNA from human HEK293T cells. However, directly impeding research on modopathies as an area of human disease is the fact that modified basecalling models for many -even the most abundant- direct RNA modifications are at their infant stages currently. In addition, cross correlation, validation with orthogonal tools, as well as diverse real-world datasets and models suitable for high-throughput sequencing are often missing. Hence the aim of this project is to train a model for pseudouridine using standard tools such as Remora, to validate on a shortlist of biologically relevant and meaningful sites and finally to cross-compare to other techniques and branch out to further modifications.