The impact of fire and herbivores on soil microbes

# **Project background**

Marie-Liesse requested to learn to run the dada2 pipeline for her existing 16S data (~250 samples). She has some previous experience with unix and using uparse and qiime (short stay at University of Pretoria).

## **Relevant info on HPC**

The pipeline used was [[ https://github.com/kviljoen/16S-rDNA-dada2-pipeline]] 

Since Marie-Liesse does not have access to the folder containing the singularity images that we (cbio group) would normally used for dada2 and fastqc, the icts_hpc.config file was edited so that the singularity images were instead pulled from quay.io repositories:

container = 'docker://quay.io/cbio/16s-rdna-dada2-pipeline'

container='docker://quay.io/h3abionet_org/h3a16s-fastqc'

This presents a better solution for future use by other institutions. In future we'd want to package all software into one image though.

We had to create a custom config file (icts_hpc.config) for Marie because our default specifies --acount cbio, whereas Marie belongs to a different group on hpc (biosci). This option will have to be included as a command line flag in future for users, along with the slurm 'queue' to be used

Another limitation is that users have to run from their home directory for now because HPC specific bind points have to be hard coded into the Dockerfile before use. It is also not enough to just include /scratch in the dockerfile with MKDIR -p as it does not pick up recursively below /scratch e.g /scratch/kviljoen/*fastq