bash loop to python loop for google colabs

Question

[svilla]

Hi there,

I have a bash loop to use kallisto on my sequencing samples, but I'm trying now to run kb on google colabs (all the fastq files are on Google drive), so I'd appreciate a hand in 'translating it to python'. Here's the original script:

#
#/bin/bash
FASTQ="/content/drive/My\ Drive/NovaSeq_raw_data/"
##############
cd $FASTQ
for FILE in $(ls *.fastq.gz | rev| cut -c 24- |rev| cut -c 14- |uniq)
do
echo "kallisto for" $FILE
kallisto quant -i $index.idx -o $FILE -b 100 "${FASTQ}WTCHG_702270_${FILE}_1.fastq.gz" "${FASTQ}WTCHG_702270_${FILE}_2.fastq.gz" "${FASTQ}WTCHG_705748_${FILE}_1.fastq.gz" "${FASTQ}WTCHG_705748_${FILE}_2.fastq.gz"
done

[Asha]

Can you help me with using colab how u downloaded SRA data to google drive?

Answer 1

[Alex Reynolds]

Untested, but perhaps it might be of use:

#!/usr/bin/env python3

import os
import subprocess

#
# redefine globals, as needed
#
index = "12345"
fastq_dir = "/content/drive/My\ Drive/NovaSeq_raw_data/"

#
# build a set of file keys, truncated per rev-cut-rev-cut-uniq pipeline
#
files = {}
for file in os.listdir(fastq_dir):
  if file.endswith(".fastq.gz"):
    file = file[13:-23]            # i.e., "rev | cut -c 24 - | rev | cut -c 14 -"
    files[file] = True             # i.e., "uniq"

#
# walk through set of file keys; call a kallisto subprocess on each 
#
for file_tuple in files.items():
  file = file_tuple[0]
  print("kallisto for {}".format(file))
  cmd = "kallisto quant -i {}.idx -o {} -b 100 {}WTCHG_702270_{}_1.fastq.gz {}WTCHG_702270_{}_2.fastq.gz {}WTCHG_705748_{}_1.fastq.gz {}WTCHG_705748_{}_2.fastq.gz".format(index, file, fastq_dir, file, fastq_dir, file, fastq_dir, file, fastq_dir, file)
  subprocess.call(cmd, shell=True)

Perhaps use print(variable_name) at various parts to see what variables look like, so that you can be sure that things are formatted properly before running the subprocess.call(...) statement.