Creating heatmaps and heattrees after haybaler

Creating graphics manually on a non-cluster R server

• Nextflow is not required for this step
• Haybaler requires an R installation
• If you don't have an R server in your cluster (quite common), you'll have to copy the haybaler_output directory to your R machine, and run the heatmaps and heattrees scripts manually.
• It's generally ok, providing, a) you've installed the haybaler dependencies correctly and b) you have R (and Rscript) installed system wide

# First transfer the haybaler_output from the nf_wochenende pipeline to your R server
# The haybaler_output should not be empty, check the CSVs first.
# Navigate to the directory
cd haybaler_output/

# Copy all the haybaler scripts to the current directory. This is an example from my setup.
cp /home/davenpor/dev/haybaler/*.sh .
cp /home/davenpor/dev/haybaler/*.py .
cp /home/davenpor/dev/haybaler/*.R .
cp -R  /home/davenpor/dev/haybaler/optional_scripts/ .
bash runbatch_heatmaps.sh

# If the system can't find Rscript, uncomment and add the path to Rscript in the runbatch_heatmaps.sh and runbatch_heattrees.sh scripts
bash runbatch_heatmaps.sh 
which Rscript
nano runbatch_heatmaps.sh

# now try again 
bash runbatch_heatmaps.sh

# activate the haybaler conda environment for the last two steps
conda activate haybaler
bash run_haybaler_tax.sh
bash runbatch_heattrees.sh

Having problems ? Please write us an issue in this repository.

Output

Output is in the directories

• top_200_taxa

• top_50_taxa

• heattree_plots