Paving the way for future sheep genome studies
Creating and analysing reference datasets on immune gene expression
Cases of liver disease are increasing worldwide, and unfortunately current treatment options remain limited. Precursors (“progenitors”) of human liver cells are a liver-specific cell type with particularly high regenerative potential. Therefore, they are a strong candidate for cell-based therapy for liver disease.
A team of researchers at the Centre for Regenerative Medicine are exploring the potential of human liver progenitor cells to regenerate. Sofia Ferreira-Gonzalez, Postdoctoral Researcher, explains that “furthering our knowledge of these cells is vital in understanding liver regeneration, and therefore effectively treating human liver conditions.”
CMVM Research Facilitator Sarah Janac met with Sofia and Alastair Kilpatrick, Bioinformatician, to discuss more about the project and the use of digital tools, especially Eddie and DataStore.
Can you tell us more about the study?
In our study, the progenitor cells were isolated from human livers that were initially intended for transplantation but were finally discarded for several reasons (for example, they may have been too fatty, or unused for logistic reasons). Even though those livers were unable to be used for transplantation, we still obtained valuable information and resources out of them, including populations of progenitor cells.
We thoroughly characterized the cells and used a bioinformatics-based approach to provide a novel transcriptomic and functional analysis. This was primarily through a type of analysis known as RNA sequencing (RNA-seq), which allows you to measure how "active" a given gene in a biological sample is. In this study, RNA-seq analysis allowed us to determine the genetic differences between healthy and diseased livers. Our analysis gave us insights into a variety of areas, including tissue repair, cell proliferation, cell specialisation and morphological characteristics.
We then transplanted these cells in our novel mouse models, which were developed to model the most important aspects of liver disease. Upon engraftment of the cells, we assessed that the progenitor cells were able to rescue the liver function, improve the liver tissue and increase the survival of these mice.
We now propose to use these hepatic progenitor cells as a first-of-its-kind cell therapy to treat human liver conditions such as biliary disease.
View publication
Let’s first talk about Eddie
At which point did you use the tool in your research, and why did you choose it?
We used the Eddie cluster for analysis of our RNA-seq data, both in the initial study and for analysis of additional samples that were requested during the peer review process. The amount of data that an RNA-seq analysis generates often runs into hundreds of gigabytes, so there are some steps in the analysis which are not possible to do without the aid of a compute cluster. The analysis was mostly run as jobs submitted to the Eddie queue, but for some tasks we used an interactive node, where we could run commands ‘in person’. We were familiar with Eddie from previous analysis projects and knew it had an extensive library of bioinformatics tools, including those required for this analysis.
Were there any challenges, and was support available?
Our DataStore was not mounted on the Eddie staging node, so this required us to transfer raw data to Eddie, and processed data back to DataStore manually, rather than use the provided staging scripts. The Eddie wiki page has a lot of useful information, and additional support was available via Information Services and the Research Computing drop-in sessions.
Do you have any advice for future users?
Check the tools available on Eddie before you start your analysis, as you may want different tools or versions. These can be installed by Information Services, or you can set up your own custom analysis environment on Eddie using a package manager such as Conda.
On to DataStore
You have already mentioned DataStore, and that it can work with Eddie. Can you tell us more about how you utilised DataStore for this project?
We used DataStore to store raw, intermediate, and processed RNA sequencing (RNA-seq) data throughout the project. Following publication of the manuscript and submission of the raw data to open-access repositories, we continue to use DataStore for storage of the processed data for use in ongoing projects within the lab.
Our RNA-seq data had originally been transferred from the sequencing facility to a local server. We transferred our data to DataStore early in the analysis process, to ensure we had a stable copy of the data with provided backups.
Were there any challenges?
Not that I remember, apart from file transfer speed at the start of the project, though this has improved recently.
Any advice for future users?
It is especially useful to have some command line expertise to make manipulation and moving data in and out easier; in this case, downloading our data from an external server and uploading to a data repository via FTP during the paper submission process.
The upper panel ("control") shows that liver progenitor cells can be cultured in vitro (i.e. outside a living organism). As they are progenitor cells, upon artificial differentiation, these cells can become other liver cells. In the lower panel ("diff"), the cells have become liver cells (hepatocytes). Hepatocytes adopt an hexagonal morphology with large centres (nuclei).
Image supplied by Dr Sofia Ferreira-Gonzalez and reproduced with permission.
Sofia Ferreira-Gonzalez, Postdoctoral Researcher, and Alastair Kilpatrick, Bioinformatician, both work at the Centre for Regenerative Medicine within the College of Medicine and Veterinary Medicine.
Read more about their research on human liver progenitor cells:
Commentary article Press release Collaboration with AstraZenecaThis case study was written by Dr Sarah Janac, Research Facilitator for the College of Medicine and Veterinary Medicine.