Topics for master theses - Hvidsten group

In my group, we use machine learning and other computational methods combined with large genomics datasets (multi-omics data) to model how genes interact in regulatory networks and how such networks give rise to properties characteristic to individuals and species. If you liked BIN315: Selected topics in functional genomics, you will love doing your master thesis work with us!

Master theses in my group start with a biological question and are typically anchored in omics data sets generated my one of my collaborators (Simen Sandve, Phil Pope, Nat Street, Siri Fjellheim or others). The focus may be on exploring a new dataset, answering specific questions using an existing data set or applying a new method to a data set.

Take some time to explore this web site and the research we do, and feel free to propose your own topics. Here are some possible starting points for master theses:

Plant gene regulatory networks

In a project funded by the Research Council of Norway (FRIPRO), we aim to understand the process of wood formation in angiosperm and gymnosperm tree species by modelling the regulatory networks orchestrating the differentiation of stem cells into woody tissues. To this end, we are currently generating multi-omics data to infer regulatory networks for each species in order to compare these networks across species to identify regulatory mechanisms explaining the evolution of trees. Collaboration with Nathaniel Street (Umeå University), Klaas Vandepoele (VIB/Ghent University), Totte Niittylä (Umeå University) and Hannele Tuominen (Umeå University).

Gene network inference through data integration

How do we infer regulatory networks from multi-omics data? Many methods have been developed, but one particularly promising agorithm is Inferelator. This method infers networks by combining expression data with transcription factor motif matches, epigenetic data (ATAC-Seq) and transcription factor binding data (DAP-seq). We would like you to help us explore how this method performs on our wood formation data!

Gene network alignment

How do we compare networks to identify rewirings that have happend during evolution of different species and how do these relate to species specific traits? Several methods exists and we would like you to help us explore how these methods perform on wood formation data from aspen and spruce!

Gene regulatory evolution after whole genome duplication

In projects funded by the Research Council of Norway (FRIPRO, Digital Life), we aim to understand how gene regulation evolves after whole genome duplication in salmonids and to reveal whether whole genome duplications spark new regulatory innovations. To this end, we are also developing phylogenetic methods for comparative transcriptomics in multiple species. Collaboration with Simen R. Sandve (BIOVIT), Rori Rohlfs (San Francisco State University) and Jon Olav Vik (BIAS/KBM).

Machine learning to reveal regulatory mechanisms

We have been combined phylogentic footprinting (transcription factor motifs obtained from various species) and open chromatin footprinting (predicting transcription factor binding using open chromatin/ATAC-seq data) to study regulatory mechanisms in Atlantic salmon. But how can we discover combinatorial regulatory patterns responsible for e.g. duplicate divergence in gene expression levels? We would like you to apply machine learning to identify such patterns!

Gene network inference through data integration

How do we infer regulatory networks from multi-omics data? Many methods have been developed, but one particularly promising agorithm is Inferelator. This method infer networks by combining expression data with transcription factor motif matches and epigenetic data (ATAC-Seq footprinting). We would like you to help us explore how this method performs on existing salmon data genereated here at NMBU!

Host-microbiota interactions

In projects funded by the Research Council of Norway and EU (Havbruk, ERANet), we are developing network-based methods for unraveling the genetic interactions between hosts and their microbial communities (holo-genomics) ... in fish and cow. Collaboration with Phil Pope and Simen R. Sandve. 

Trans-kingdom host-microbiota interactions

A former master student, Marius Strand, developed a method for inferring trans-kingdom host-microbiota networks and associating these to external variabels such as feed type in Atlantic salmon. We would like you to develop this method further and apply it to new data!

Other collaborative projects

Developing an analysis workflow for the analysis of multi-OMICS data

The goal of this master thesis project is to develop a pipeline for the analysis and integration of multi-OMICS data that originate from ecotoxicological experiments with endocrine disrupting chemicals in the water flea Daphnia magna. The analysis workflow will be established using an existing dataset and will find its application in the analysis of a data set of experiments conducted with chitin synthesis inhibitors in D. magna. Collaboration with NIVA.

Previous master theses

1. Marius Strand. Exploring host-microbiome interactions in Norwegian Salmon via weighted network analysis. Master thesis, NMBU, 2019.
2. Cathrine Horntvedt Kristiansen. The effect of chromatin structure on duplicate gene expression in Atlantic salmon. Master thesis, NMBU, 2019.
3. Eivind Kjeka Broen. Comparison between gene expression and protein abundance in Populus tremula wood development. Master thesis, NMBU, 2019.
4. Katrine Hånes Kirste. Optimizing Transcriptome Analysis using Short-Read RNA-Seq in Atlantic Salmon. Master thesis, NMBU, 2016.
   
5. Ine Birgitta Hallsberg. Homeolog Regulation in Hexaploid Wheat. Master thesis, NMBU, 2016.
   
6. Yonatan Ayalew Mekonnen. Evaluation of GWAS Method Performance Focusing on Population Stratification and Cryptic Relatedness. Master thesis, NMBU, 2015.
   
7. Xiao Nie. Filtering the co-expression networks of populus trichocarpa. Master thesis, NMBU, 2014.
   
8. Jonas Christoffer Lindstrøm. Comparative analysis of plant regulatory genomes. Master thesis, NMBU, 2014.
   
9. Niklas Mähler. Computational prediction of gene regulation in Synechocystis sp. PCC6803. Master thesis, UmU, 2012. Paper in PLoS ONE.
   
10. Patrik Björkholm. Method for recognizing local descriptors of protein structures using Hidden Markov Models, Master thesis, UU, 2008. Paper in Bioinformatics.
    
11. Minyan Hong. Fold recognition using local descriptors of protein structure and Hidden Markov Models, Student project (10 points), 2006 and Master thesis, UU, 2007.
    
12. Marta Luksza. A System for Predicting Protein Function from Structure, Master thesis, UU, 2005.