There is significant consensus that the alteration of tissue
microenvironment in pancreatic tissue (e.g., infiltration of
immune cells, β-cell death, fibrosis) is involved in several
pancreatic pathologies.
My current project aims to set up a computational
infrastructure to analyze spatial transcriptomics data of
pancreatic tissue. I will optimize the computational
methodology to study spatial transcriptomics data from
pancreatic biopsies from healthy donors and donors
with different types of pancreatic diseases. To do this, I
will, among other things, generate spatial transcriptomics
data and run FASTQ files through Snakemake/Nextflow
pipelines via Singularity. I will also use different
data analysis and exploration tools such as Cell2Deconvolution.
In this project, I am in need of storage and computational
resources to store and run my pipelines and data to
perform proper evaluation. From the usage distribution of
the previous year, I realized that I used more resources than
I had planned for in my initial proposal: ‘My current first
sample takes about 500 CPU hours, and I will analyze
about 25 of these samples (totalling approximately 25x500
CPU hours). To also allow time for troubleshooting and
interactive sessions, I require an additional 5x1000
core-hours. I have previously worked on my local
computer and using cloud services, but I do not see this
strategy as sustainable anymore.
As I, in the previous round, used approximately a total of
165 X 1000 core-h. Hence, I will use this as a motivation
behind asking for the maximum number of core-h hours
available.
Classification: Bioinformatics, Computational Biology,
Transcriptomics, Medicine, Biomedical Science