DATA_OUT: GC bias plot (picard -> all plex on one plot)

Cleiti / UpSPLAT

0 stars 0 forks source link

Closed Cleiti closed 1 year ago

Cleiti commented 1 year ago

Input: cram data, here: run_sarek/results/preprocessing/markduplicates (search for cram (and crai?) using make_fastqlist.sh)

Module: module load bioinfo-tools module load picard/2.0.1

call: java -Xmx16g -jar $PICARD_HOME/picard.jar CollectGcBiasMetrics option1=opt opt2=opt

Cleiti commented 1 year ago

Result: bash file called makeGcPlot.sh

Bug to solve:

Cleiti commented 1 year ago

OBS: requires reference genome -> sample-dependent! (Right now has reference genome for bigYeast)

Cleiti commented 1 year ago

TODO:

slurm it (OBS still one thread, picard has no option)
create another script to plot all normalized coverage as line in same plot for each barcode

Cleiti commented 1 year ago

what created GCbias folder?

[ ] make gc_figure.R : sarek_results/GCbias/-barcode-/gc_biasmetrics.txt -> plot normalized coverage as 1 line for the barcode
[ ] write system to go through the GCbias folder for all barcodes and plot one png with 48 lines.

Suggestion (meh???) on system to make a plot with all data:

Cleiti commented 1 year ago

UPD:

make_gc_bias runs picard and generates DATA in sarek_results/GCbias

now need: parse txt plot 1 (already have code somwehere?) parse all txt plot all on one + legend