Adding cell_ranger _loop python modification

Single_cell_RNAseq_10x/cellranger_loop_pymod.sh

@@ -0,0 +1,129 @@
+#!/bin/bash
+#BSUB cwd /gpfs/projects/cek26
+#BSUB -J cellranger_loop
+#BSUB -q bsc_debug
+#BSUB -W 1:00
+#BSUB -eo /gpfs/projects/cek26/SANTAMARIA_09_TMP/bsub_reports/cellranger_loop.err
+#BSUB -oo /gpfs/projects/cek26/SANTAMARIA_09_TMP/bsub_reports/cellranger_loop.out
+#BSUB -M 1800
+#BSUB -n 1
+
+
+# Load modules in order to run cellranger
+module load singularity 
+module load PYTHON/3.9.0
+
+# Set and switch to your working directory
+wd=/gpfs/projects/cek26/SANTAMARIA_09_TMP
+cd $wd
+
+# Set reference directory
+ref=/gpfs/projects/cek26/genome_mus/cellranger/refdata-gex-mm10-2020-A
+
+# Scripts directory
+scr=/gpfs/projects/cek26/software/scripts
+
+for f in $(find fastq_files -name  "*.gz")
+do
+echo $f
+gzip -t $f && echo ok || echo corrupted
+done
+
+#Rename fastq files to cellranger input format
+#Remember to setup the correct paths for the arguments
+#To return help about the usage run: fastq_cellranger_format_converter.py -h
+
+for f in $(find fastq_files -name  "*.gz")
+do
+echo $f
+gzip -t $f && echo ok || echo corrupted
+done
+python $scr/fastq_cellranger_format_converter.py -t $wd/SANTAMARIA_09.txt -fd $wd/fastq_files -r False -cwd $wd
+#Check if the files are renamed propperly and all expected are present
+sort $wd/fastq_comp1.txt > $wd/sorted_fastq_comp1.txt
+sort $wd/fastq_comp2.txt > $wd/sorted_fastq_comp2.txt
+cmp $wd/sorted_fastq_comp1.txt $wd/sorted_fastq_comp2.txt || echo "Fastq files missing!"
+
+
+
+# Create folders results and store FASTQfiles in a new folder [/fastq_files] inside the working directory, use the correct name nomenclature!
+# [Sample Name]_S1_L00[Lane Number]_[Read Type]_001.fastq.gz
+# Where Read Type is one of:
+# I1: Sample index read (optional)
+# R1: Read 1
+# R2: Read 2
+mkdir cellranger_counts
+mkdir to_bsub
+
+# Loop to generate cellranger counts for each sample
+# Firstly, generate a list of the samples you have
+for f in $(find ./fastq_files -name "*.fastq.gz" -exec basename {} \;)
+do
+echo $(cut -d'_' -f1 <<< $f) >> samples.txt
+done
+
+# Then use a loop with the unique samples to generate respective jobs to be sent to bsub
+# if expected number of cells differ, the jobs need to be changed manually
+for f in $(sort -u samples.txt)
+do
+{
+echo \#!/bin/bash
+echo \#BSUB cwd $wd
+echo \#BSUB -J cellranger_counts_$f
+echo \#BSUB -q bsc_ls
+echo \#BSUB -W 42:00
+echo \#BSUB -eo $wd/bsub_reports/cellranger_counts_$f.err
+echo \#BSUB -oo $wd/bsub_reports/cellranger_counts_$f.out
+echo \#BSUB -M 3000
+echo \#BSUB -n 256
+echo \#BSUB "span[ptile=16]"
+echo \#BSUB -x
+
+echo module load singularity
+echo cd $wd/cellranger_counts
+echo singularity exec /apps/CELLRANGER/SRC/images/cellranger-6.0.1.sif cellranger count --id=$f \
+                 --transcriptome=$ref \
+                 --fastqs=$wd/fastq_files \
+                 --expect-cells=10000 \
+                 --sample=$f
+} > to_bsub/cellranger_count_$f.sh
+sed -i -e 's/\r$//' to_bsub/cellranger_count_$f.sh
+bsub < to_bsub/cellranger_count_$f.sh
+done
+
+
+# Next steps will be on hold until previous pipelines are over
+while [ $(sort -u samples.txt | wc -l) != $(find ./cellranger_counts -type f -name "*_info.h5" -print | wc -l) ]
+do
+sleep 300;
+done
+
+# Generate aggregation CSV 
+for i in $(sort -u samples.txt)
+do
+echo $(cut -d'_' -f1 <<< $i) >> experiment.txt
+done
+aggr_id=()
+for u in $(sort -u experiment.txt)
+do
+aggr_id=$aggr_id'_'$u
+done
+aggr_id=Aggr$aggr_id
+echo library_id,molecule_h5 > $aggr_id.csv
+for i in $(sort -u samples.txt)
+do
+echo $i,$wd/cellranger_counts/$i/outs/molecule_info.h5 >> $aggr_id.csv
+done
+
+
+# Run cellranger aggr when counts finish
+mkdir cellranger_aggr
+cd cellranger_aggr
+singularity exec /apps/CELLRANGER/SRC/images/cellranger-6.0.1.sif cellranger aggr --id=$aggr_id \
+                --csv=$wd/$aggr_id.csv \
+                --normalize=mapped
+
+
+# Remove used scripts
+rm -r ../to_bsub
+rm ../experiment.txt