Hello!
We’re back from winter vacation and it’s time for a new GenomeDK newsletter.
Enjoy!
The GenomeDK Team
More compute coming in 2020
In 2019 we said goodbye to our s01 and s02 nodes. These were the original nodes from the purchase that kickstarted GenomeDK in 2012. The nodes were out of warranty and started to cause issues since they were connected to the cluster through a different network than the remaining nodes, which caused a bottleneck that affected all users.
Taking these nodes off the grid has caused a small queue to build up since we now have fewer compute nodes available. However, we’re currently in the final stages of submitting a tender for new compute nodes. Hopefully this will result in a significant increase of our compute power in 2020.
Workshops
We cover topics like general cluster usage, gwf workflows, software management with Conda, and how to organize your projects. If your group/center/institute would like a workshop, don’t hesitate to contact us. You can read more here.
We also hosted a very popular official workshop in December. Next time we’ll make sure to book a larger room and get more cake!
Stay up to date on your compute and storage usage
The space command has been updated. It now has different views for users, projects, and project owners. Additionally, compute usage over time is now also reported. You can read more about the space commands here and here.
New storage has been completely attached
Since the last newsletter the new expansion for faststorage has been completed. This included moving several petabytes of project data from old to new storage and then reattaching the old storage. This process was completely transparent to users.
Old project folders are being moved to faststorage
Old projects were created on our NFS filesystem. NFS is not suitable for the kind of data that projects usually contain, so we’re now in the process of moving those old projects to fast storage. To do this we need regular downtimes that have been (and still will be) announced on the status page.
Lucie was one of the most active users on GenomeDK in 2019. We asked her to say a bit about what she’s been doing.
I am Lucie Bergeron and I am currently enrolled as a PhD Fellow at Copenhagen University under the supervision of Guojie Zhang (http://zhanggjlab.org/en/).
I started my project on germline mutation rate in vertebrates 2 years ago, yes, it goes beyond mammals! During the first year of my PhD I collected trio samples (parents and offspring) for 75 species of vertebrates including 44 mammals, 17 birds, 8 fishes and 6 reptiles.
We sequenced each individual at 60X coverage and then we will map the reads and call variants. Finally, by comparing the genome of the parents to the one of their offspring, we can find the number of germline mutation that occurs in one generation. For most species, I have 1 or 2 trios but for few I have more than 5 trios. Especially, I’ve got a big pedigree of rhesus macaque (Macaca mulatta) with 19 trios.
With this dataset I developed my pipeline and got some interesting results that we could compare to humans and apes mutational processes.
My project is in collaboration with Søren Besenbacher and Mikkel H. Schierup from Aarhus University who introduced me to GenomeDK! Since then I am using it extensively to analyze all my genomes.
Thanks to Lucie for sharing her story with us!
Thank you for reading!