Mercurial > repos > devteam > tabular_to_fastq

annotate tabular_to_fastq.py @ 4:2dcfbbf9071a draft

Find changesets by keywords (author, files, the commit message), revision number or hash, or revset expression.
"planemo upload for repository https://github.com/galaxyproject/tools-iuc/tree/master/tool_collections/galaxy_sequence_utils/tabular_to_fastq commit d4ced60a941c4c4a2fe95de9c09a10086810b387"
author iuc
date Wed, 19 Feb 2020 12:59:28 -0500
parents b8cdc0507586
children
Ignore whitespace changes - Everywhere: Within whitespace: At end of lines:
rev   line source
2
b8cdc0507586 planemo upload for repository https://github.com/galaxyproject/tools-devteam/tree/master/tool_collections/galaxy_sequence_utils/tabular_to_fastq commit f2582539542b33240234e8ea6093e25d0aee9b6a
devteam
parents: 0
diff changeset
1 # Dan Blankenberg
b8cdc0507586 planemo upload for repository https://github.com/galaxyproject/tools-devteam/tree/master/tool_collections/galaxy_sequence_utils/tabular_to_fastq commit f2582539542b33240234e8ea6093e25d0aee9b6a
devteam
parents: 0
diff changeset
2 from __future__ import print_function
b8cdc0507586 planemo upload for repository https://github.com/galaxyproject/tools-devteam/tree/master/tool_collections/galaxy_sequence_utils/tabular_to_fastq commit f2582539542b33240234e8ea6093e25d0aee9b6a
devteam
parents: 0
diff changeset
3
0
b334cd1095ea Imported from capsule None
devteam
parents:
diff changeset
4 import sys
b334cd1095ea Imported from capsule None
devteam
parents:
diff changeset
5
2
b8cdc0507586 planemo upload for repository https://github.com/galaxyproject/tools-devteam/tree/master/tool_collections/galaxy_sequence_utils/tabular_to_fastq commit f2582539542b33240234e8ea6093e25d0aee9b6a
devteam
parents: 0
diff changeset
6
0
b334cd1095ea Imported from capsule None
devteam
parents:
diff changeset
7 def main():
b334cd1095ea Imported from capsule None
devteam
parents:
diff changeset
8 input_filename = sys.argv[1]
b334cd1095ea Imported from capsule None
devteam
parents:
diff changeset
9 output_filename = sys.argv[2]
2
b8cdc0507586 planemo upload for repository https://github.com/galaxyproject/tools-devteam/tree/master/tool_collections/galaxy_sequence_utils/tabular_to_fastq commit f2582539542b33240234e8ea6093e25d0aee9b6a
devteam
parents: 0
diff changeset
10 identifier_col = int(sys.argv[3]) - 1
b8cdc0507586 planemo upload for repository https://github.com/galaxyproject/tools-devteam/tree/master/tool_collections/galaxy_sequence_utils/tabular_to_fastq commit f2582539542b33240234e8ea6093e25d0aee9b6a
devteam
parents: 0
diff changeset
11 sequence_col = int(sys.argv[4]) - 1
b8cdc0507586 planemo upload for repository https://github.com/galaxyproject/tools-devteam/tree/master/tool_collections/galaxy_sequence_utils/tabular_to_fastq commit f2582539542b33240234e8ea6093e25d0aee9b6a
devteam
parents: 0
diff changeset
12 quality_col = int(sys.argv[5]) - 1
b8cdc0507586 planemo upload for repository https://github.com/galaxyproject/tools-devteam/tree/master/tool_collections/galaxy_sequence_utils/tabular_to_fastq commit f2582539542b33240234e8ea6093e25d0aee9b6a
devteam
parents: 0
diff changeset
13
b8cdc0507586 planemo upload for repository https://github.com/galaxyproject/tools-devteam/tree/master/tool_collections/galaxy_sequence_utils/tabular_to_fastq commit f2582539542b33240234e8ea6093e25d0aee9b6a
devteam
parents: 0
diff changeset
14 max_col = max(identifier_col, sequence_col, quality_col)
0
b334cd1095ea Imported from capsule None
devteam
parents:
diff changeset
15 num_reads = None
b334cd1095ea Imported from capsule None
devteam
parents:
diff changeset
16 skipped_lines = 0
4
2dcfbbf9071a "planemo upload for repository https://github.com/galaxyproject/tools-iuc/tree/master/tool_collections/galaxy_sequence_utils/tabular_to_fastq commit d4ced60a941c4c4a2fe95de9c09a10086810b387"
iuc
parents: 2
diff changeset
17 with open(output_filename, 'w') as out:
2dcfbbf9071a "planemo upload for repository https://github.com/galaxyproject/tools-iuc/tree/master/tool_collections/galaxy_sequence_utils/tabular_to_fastq commit d4ced60a941c4c4a2fe95de9c09a10086810b387"
iuc
parents: 2
diff changeset
18 for num_reads, line in enumerate(open(input_filename)):
2dcfbbf9071a "planemo upload for repository https://github.com/galaxyproject/tools-iuc/tree/master/tool_collections/galaxy_sequence_utils/tabular_to_fastq commit d4ced60a941c4c4a2fe95de9c09a10086810b387"
iuc
parents: 2
diff changeset
19 fields = line.rstrip('\n\r').split('\t')
2dcfbbf9071a "planemo upload for repository https://github.com/galaxyproject/tools-iuc/tree/master/tool_collections/galaxy_sequence_utils/tabular_to_fastq commit d4ced60a941c4c4a2fe95de9c09a10086810b387"
iuc
parents: 2
diff changeset
20 if len(fields) > max_col:
2dcfbbf9071a "planemo upload for repository https://github.com/galaxyproject/tools-iuc/tree/master/tool_collections/galaxy_sequence_utils/tabular_to_fastq commit d4ced60a941c4c4a2fe95de9c09a10086810b387"
iuc
parents: 2
diff changeset
21 out.write("@%s\n%s\n+\n%s\n" % (fields[identifier_col], fields[sequence_col], fields[quality_col]))
2dcfbbf9071a "planemo upload for repository https://github.com/galaxyproject/tools-iuc/tree/master/tool_collections/galaxy_sequence_utils/tabular_to_fastq commit d4ced60a941c4c4a2fe95de9c09a10086810b387"
iuc
parents: 2
diff changeset
22 else:
2dcfbbf9071a "planemo upload for repository https://github.com/galaxyproject/tools-iuc/tree/master/tool_collections/galaxy_sequence_utils/tabular_to_fastq commit d4ced60a941c4c4a2fe95de9c09a10086810b387"
iuc
parents: 2
diff changeset
23 skipped_lines += 1
2
b8cdc0507586 planemo upload for repository https://github.com/galaxyproject/tools-devteam/tree/master/tool_collections/galaxy_sequence_utils/tabular_to_fastq commit f2582539542b33240234e8ea6093e25d0aee9b6a
devteam
parents: 0
diff changeset
24
0
b334cd1095ea Imported from capsule None
devteam
parents:
diff changeset
25 if num_reads is None:
2
b8cdc0507586 planemo upload for repository https://github.com/galaxyproject/tools-devteam/tree/master/tool_collections/galaxy_sequence_utils/tabular_to_fastq commit f2582539542b33240234e8ea6093e25d0aee9b6a
devteam
parents: 0
diff changeset
26 print("Input was empty.")
0
b334cd1095ea Imported from capsule None
devteam
parents:
diff changeset
27 else:
2
b8cdc0507586 planemo upload for repository https://github.com/galaxyproject/tools-devteam/tree/master/tool_collections/galaxy_sequence_utils/tabular_to_fastq commit f2582539542b33240234e8ea6093e25d0aee9b6a
devteam
parents: 0
diff changeset
28 print("%i tabular lines were written as FASTQ reads. Be sure to use the FASTQ Groomer tool on this output before further analysis." % (num_reads + 1 - skipped_lines))
b8cdc0507586 planemo upload for repository https://github.com/galaxyproject/tools-devteam/tree/master/tool_collections/galaxy_sequence_utils/tabular_to_fastq commit f2582539542b33240234e8ea6093e25d0aee9b6a
devteam
parents: 0
diff changeset
29
b8cdc0507586 planemo upload for repository https://github.com/galaxyproject/tools-devteam/tree/master/tool_collections/galaxy_sequence_utils/tabular_to_fastq commit f2582539542b33240234e8ea6093e25d0aee9b6a
devteam
parents: 0
diff changeset
30
b8cdc0507586 planemo upload for repository https://github.com/galaxyproject/tools-devteam/tree/master/tool_collections/galaxy_sequence_utils/tabular_to_fastq commit f2582539542b33240234e8ea6093e25d0aee9b6a
devteam
parents: 0
diff changeset
31 if __name__ == "__main__":
b8cdc0507586 planemo upload for repository https://github.com/galaxyproject/tools-devteam/tree/master/tool_collections/galaxy_sequence_utils/tabular_to_fastq commit f2582539542b33240234e8ea6093e25d0aee9b6a
devteam
parents: 0
diff changeset
32 main()