alevin-fry
SCRIPT_DIR=$(pwd)
OUT_DIR='./results'
CELLRANGER_OUTPUTS='/home/placeholder/bam_dir'
SALMON_INDEX='./salmon_index/'
tgMap_FILE='tg2.txt'
CELLRANGER_BARCODES=/apps/prod/easybuild/sl7.x86_64.foss-2021a/software/cellranger/7.0.0/lib/python/cellranger/barcodes/737K-august-2016.txt
mkdir -p ${OUT_DIR}
# Input list with name of all samples to run scHLAtype on
input_list='list_of_samples.txt'
readarray -t array_list < ${input_list}
SAMPLE_NAME=${array_list[$SLURM_ARRAY_TASK_ID-1]}
# Output folder for each sample
OUT_DIR_SAMPLE=${OUT_DIR}/${SAMPLE_NAME}
# Path to BAM file that was created by cellranger multi
ORIGINAL_BAM_FILE=${CELLRANGER_OUTPUTS}/${SAMPLE_NAME}/outs/possorted_genome_bam.bam
echo "check1"
# Extract fastq files
cellranger bamtofastq --nthreads=16 --reads-per-fastq=9223372036854775807 ${ORIGINAL_BAM_FILE} ${OUT_DIR_SAMPLE}
# Check header to know which library contains Gene Expression
regex_expr='"library_id":([0-9])'
samtools_header=$(samtools view -H ${ORIGINAL_BAM_FILE} | grep "Gene Expression")
echo ${samtools_header}
# Do actual regex to know which library contains Gene Expression
[[ $samtools_header =~ $regex_expr ]]
library_number=${BASH_REMATCH[1]}
cd ${OUT_DIR_SAMPLE}
echo "check2"
combine_grep=_${library_number}_1_
gex_folder=$(realpath $(ls | grep "${combine_grep}"))
echo Identified ${gex_folder} as correct Gene Expression library
cd ${gex_folder}
# Get fastq files
FASTQ_ONE=$(echo $(ls -1 -d "$PWD/"* | grep "_R1_") | tr -s ' ')
FASTQ_TWO=$(echo $(ls -1 -d "$PWD/"* | grep "_R2_") | tr -s ' ')
cd ${SCRIPT_DIR}

chmod -R 770 ~/salmon/results
echo "check3"

# Run salmon alevin
salmon alevin -l ISF -1 ${FASTQ_ONE} -2 ${FASTQ_TWO} --chromium -i ${SALMON_INDEX} -p 16 -o ${OUT_DIR_SAMPLE}/salmon --tgMap ${tgMap_FILE} --sketch
# Run alevin fry
# generate-permit-list: Now, we can use this permit list to scan the cell barcodes actually encountered in our
#reads and determine a set of cells that were likely present in our sample
echo "check4"
alevin-fry generate-permit-list --input ${OUT_DIR_SAMPLE}/salmon --expected-ori rc --output-dir ${OUT_DIR_SAMPLE}/salmon/alevin_fry --unfiltered-pl ${CELLRANGER_BARCODES}
echo "check5"
# collate: Given the permit list and barcode mapping (which resides in the <fry_odir> directory), we collate
#the original RAD file using the command below.
alevin-fry collate -i ${OUT_DIR_SAMPLE}/salmon/alevin_fry --rad-dir ${OUT_DIR_SAMPLE}/salmon/ -t 16
echo "check6"
# quant: Finally, we quantify the collated rad file using the cr-like-em resolution strategy using the quant
#command below.
alevin-fry quant -i ${OUT_DIR_SAMPLE}/salmon/alevin_fry -m ${tgMap_FILE} --resolution cr-like-em -o ${OUT_DIR_SAMPLE}/salmon/alevin_fry --dump-eqclasses --use-mtx -t 16
echo "check7"
# Run infer
alevin-fry infer --count-mat ${OUT_DIR_SAMPLE}/salmon/alevin_fry/alevin/geqc_counts.mtx --eq-labels ${OUT_DIR_SAMPLE}/salmon/alevin_fry/alevin/gene_eqclass.txt.gz --output-dir ${OUT_DIR_SAMPLE}/salmon/alevin_fry_final -t 16