Assistant Professor

Work mail: jakob.nissen@sund.ku.dk

Private mail: jakobnybonissen@gmail.com

Photo of Jakob Nissen

Short profile

I was trained as a molecular biologist, before pivoting to bioinformatics during my PhD. Currently I work as somewhere in between a researcher and a research software engineer.

I’m enthusiastic about scientific programming, molecular evolution, and teaching.

Current projects

  • PlasMAAG: Creation of a plasmid binner from short-read metagenomic whole genome sequencing data
  • TaxVamb: Creation of taxonomy guided semi-supervised metagenomic binner for organisms.

Research interests

  • Metagenomics and microbial bioinformatics
  • Computational molecular evolution and phylogenetics
  • Research software engineering
  • Sub-genomic entiries, viruses, plasmids, transposons

Future projects (open for collaborations):

Amanuensis: Creation of a new read simulator for highly realistic HiFi reads.

There exists lots of programs for simulating sequencing reads, but in my experience, none of them realistically mimic modern Pacific BioSciences HiFi reads. Being able to create synthetic HiFi datasets is crucial to validate methods in metagenomics.

Simulated dataset: Creation of new simulated metagenomic datasets for validation of metagenomic software

Most new methods in metagenomics are validated using simulated data. Unfortunately, the current simulation state of the art is outdated, and has severe inaccuracies. We need better simulated datasets in order for the field to progress

Supervised binner: Hypothetical future metagenomic binner using constrained clustering constrained by a supervised deep learning model

Almost all modern binners uses a paradigm where the input is embedded using a neural network, and this embedding is clustered. However, we don’t have good ways of training the network to produce this embedding. A possible new paradigm would be to train neural networks using the simpler, old-fashioned supervised manner to output constraints for the clustering, and then cluster with a non-deep learning clustering algorithm.

Datasets

  • Critical Assessment of Metagenome Interpretation, rounds I and II

Short CV

  • 2022 - present: Assistant professor at Novo Nordisk Foundation Center for Basic Metabolic Research, University of Copenhagen, Denmark.
  • 2020-2022: Post doc at the Influenza section, Statens Serum Institut.
  • 2016-2020: PhD at section of Bioinformatics, Department of Health Technology, Technical University of Denmark.
  • 2014-2016: MSc in Omics, University of Southern Denmark / University of Chinese Academy of Sciences.
  • 2011-2014: Bsc in biology, University of Copenhagen

GitHub - jakobnissen