PhD Projects » ESR 5: Bacteriophage–host relationships prediction

Xue Peng

I am Xue Peng and come from China. I obtained my bachelor and master degree both from Sun Yat-sen University, China.

My master major is ecology. During my study, I am fascinated by the complex ecosystem with several layers of interaction within the microbes. After graduation, I join BGI Genomics as a bioinformatics engineer in the metagenomic research group. I am responsible for developing metagenomic analysis tools and doing some research works about how gut bacteria affect human health. Recently we used mathematic methods to pick 5635 reference genomes database to represent human gut bacteria and archaea species. This database will provide a higher resolution about human gut bacteria identification.

This project is amazing. I can take a deep look at the interaction within the microbes, such as how phage finds its bacteria partner and how they functionally linked to each other. It is a great chance to expand my knowledge in virology, and I believe we can find interesting stuffs in virology.

Also, I am interested in the renaissance of Hanfu culture. I want the whole world can appreciate the beauty of Chinese Hanfu. Other interest, photography, reading and puzzle adventure game.

Host institution:
Helmholtz Zentrum München (HMGU), Germany
Local supervisor:
Dr. Li Deng (HMGU)
Local co-supervisor:
Prof. Dr. Caroline Friedel (Ludwig Maximilians University Munich)
Project partner:
ESR 10
Work packages:
WP 1.1 Virus identification
WP 1.2 Host prediction
WP 2.1 Microevolution: Virus quasispecies

Li Deng
Caroline Friedel

Project description

This project seeks to understand how (bacterio-)phages are functionally linked to their Bacteria or Archaea hosts both using experimental and bioinformatics approach. Phage identified using cultivation-independent viral metagenomic sequencing lack sequence similarity to any of previously sequenced ones can hardly be assigned to their hosts. Different methods exist to predict hosts for phage sequence, e.g. single or multiple features (co-abundance, sequence homology, similarity to other phages or sequence composition similarity between phages and their hosts) or using these features in machine learning models.

Therefore, we aim to integrate features of our unique experimental identified, host-linked viral metagenomic data in a deep learning model to improve the prediction performance.

  • Development of a modified version for experimentally linking phages to their anaerobic bacteria and archaea hosts based on the previously published culture-independent viral-tagging (VT) method for aerobic bacteria-phage pairs. In this single cell viral tagging approach, virions from an environmental sample are fluorescently labelled through DNA staining and upon them infecting host cells, the infected host is labelled. This is followed by single cell sorting via FACS of the infected hosts together with the phages infecting. We plan to use three bacterial strains (complete genome available) to VT the same 20 water samples to create 60 VT metagenomic datasets, as well as 20 community viral metagenomics datasets.
  • Metagenomic sequencing of the specific phage communities on selected bacteria and archaea hosts as both bulk level and single cell level.
  • Single phage-host screening, isolation and sequencing based on selected gene markers representing different adsorption or infecting strategies (together with ESR 10).
  • Study the adsorption and host recognition mechanisms of the newly isolated phages.
  • Generation of viral-tagged metagenomes on the same selected bacteria and archaea hosts in contaminated groundwater ecosystems, and community co-occurrence microbiomes and viromes in the same ecosystems (together with ESR 10 and BioRelate).

Doctorate will be awarded at Technical University of Munich.