Mercurial > repos > iuc > trycycler_subsample

annotate trycycler.py @ 6:392c8f5d213b draft default tip

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/tools/trycycler commit 9da13f57f93e69fd463f6245af45674fe011b861
author iuc
date Wed, 06 Nov 2024 13:29:01 +0000
parents deb7b00e877e
children
Ignore whitespace changes - Everywhere: Within whitespace: At end of lines:
rev   line source
0
deb7b00e877e "planemo upload for repository https://github.com/galaxyproject/tools-iuc/tree/master/tools/trycycler commit 9d7c4277b0f96aacd466f2d497e08edcca3fa238"
iuc
parents:
diff changeset
1 #!/usr/bin/env python3
deb7b00e877e "planemo upload for repository https://github.com/galaxyproject/tools-iuc/tree/master/tools/trycycler commit 9d7c4277b0f96aacd466f2d497e08edcca3fa238"
iuc
parents:
diff changeset
2
deb7b00e877e "planemo upload for repository https://github.com/galaxyproject/tools-iuc/tree/master/tools/trycycler commit 9d7c4277b0f96aacd466f2d497e08edcca3fa238"
iuc
parents:
diff changeset
3 from os import path, walk
deb7b00e877e "planemo upload for repository https://github.com/galaxyproject/tools-iuc/tree/master/tools/trycycler commit 9d7c4277b0f96aacd466f2d497e08edcca3fa238"
iuc
parents:
diff changeset
4 from sys import argv
deb7b00e877e "planemo upload for repository https://github.com/galaxyproject/tools-iuc/tree/master/tools/trycycler commit 9d7c4277b0f96aacd466f2d497e08edcca3fa238"
iuc
parents:
diff changeset
5
deb7b00e877e "planemo upload for repository https://github.com/galaxyproject/tools-iuc/tree/master/tools/trycycler commit 9d7c4277b0f96aacd466f2d497e08edcca3fa238"
iuc
parents:
diff changeset
6
deb7b00e877e "planemo upload for repository https://github.com/galaxyproject/tools-iuc/tree/master/tools/trycycler commit 9d7c4277b0f96aacd466f2d497e08edcca3fa238"
iuc
parents:
diff changeset
7 def cluster(output_folder):
deb7b00e877e "planemo upload for repository https://github.com/galaxyproject/tools-iuc/tree/master/tools/trycycler commit 9d7c4277b0f96aacd466f2d497e08edcca3fa238"
iuc
parents:
diff changeset
8 counter = 1
deb7b00e877e "planemo upload for repository https://github.com/galaxyproject/tools-iuc/tree/master/tools/trycycler commit 9d7c4277b0f96aacd466f2d497e08edcca3fa238"
iuc
parents:
diff changeset
9 for root, dir, files in walk(output_folder):
deb7b00e877e "planemo upload for repository https://github.com/galaxyproject/tools-iuc/tree/master/tools/trycycler commit 9d7c4277b0f96aacd466f2d497e08edcca3fa238"
iuc
parents:
diff changeset
10 if root.endswith('1_contigs'):
deb7b00e877e "planemo upload for repository https://github.com/galaxyproject/tools-iuc/tree/master/tools/trycycler commit 9d7c4277b0f96aacd466f2d497e08edcca3fa238"
iuc
parents:
diff changeset
11 output_path = path.join(output_folder, f"cluster_0{counter}.fasta")
deb7b00e877e "planemo upload for repository https://github.com/galaxyproject/tools-iuc/tree/master/tools/trycycler commit 9d7c4277b0f96aacd466f2d497e08edcca3fa238"
iuc
parents:
diff changeset
12 with open(output_path, "a") as out_cluster:
deb7b00e877e "planemo upload for repository https://github.com/galaxyproject/tools-iuc/tree/master/tools/trycycler commit 9d7c4277b0f96aacd466f2d497e08edcca3fa238"
iuc
parents:
diff changeset
13 for fasta in files:
deb7b00e877e "planemo upload for repository https://github.com/galaxyproject/tools-iuc/tree/master/tools/trycycler commit 9d7c4277b0f96aacd466f2d497e08edcca3fa238"
iuc
parents:
diff changeset
14 fasta_path = path.join(root, fasta)
deb7b00e877e "planemo upload for repository https://github.com/galaxyproject/tools-iuc/tree/master/tools/trycycler commit 9d7c4277b0f96aacd466f2d497e08edcca3fa238"
iuc
parents:
diff changeset
15 fasta = open(fasta_path).read()
deb7b00e877e "planemo upload for repository https://github.com/galaxyproject/tools-iuc/tree/master/tools/trycycler commit 9d7c4277b0f96aacd466f2d497e08edcca3fa238"
iuc
parents:
diff changeset
16 out_cluster.write(fasta)
deb7b00e877e "planemo upload for repository https://github.com/galaxyproject/tools-iuc/tree/master/tools/trycycler commit 9d7c4277b0f96aacd466f2d497e08edcca3fa238"
iuc
parents:
diff changeset
17 counter += 1
deb7b00e877e "planemo upload for repository https://github.com/galaxyproject/tools-iuc/tree/master/tools/trycycler commit 9d7c4277b0f96aacd466f2d497e08edcca3fa238"
iuc
parents:
diff changeset
18
deb7b00e877e "planemo upload for repository https://github.com/galaxyproject/tools-iuc/tree/master/tools/trycycler commit 9d7c4277b0f96aacd466f2d497e08edcca3fa238"
iuc
parents:
diff changeset
19
deb7b00e877e "planemo upload for repository https://github.com/galaxyproject/tools-iuc/tree/master/tools/trycycler commit 9d7c4277b0f96aacd466f2d497e08edcca3fa238"
iuc
parents:
diff changeset
20 def reconcile(input_file):
deb7b00e877e "planemo upload for repository https://github.com/galaxyproject/tools-iuc/tree/master/tools/trycycler commit 9d7c4277b0f96aacd466f2d497e08edcca3fa238"
iuc
parents:
diff changeset
21 number_cluster = [x for x in input_file[-1:0:-1] if x.isdigit()][0]
deb7b00e877e "planemo upload for repository https://github.com/galaxyproject/tools-iuc/tree/master/tools/trycycler commit 9d7c4277b0f96aacd466f2d497e08edcca3fa238"
iuc
parents:
diff changeset
22 full_path = f"selected_cluster/cluster_0{number_cluster}/1_contigs/"
deb7b00e877e "planemo upload for repository https://github.com/galaxyproject/tools-iuc/tree/master/tools/trycycler commit 9d7c4277b0f96aacd466f2d497e08edcca3fa238"
iuc
parents:
diff changeset
23 with open(input_file) as tmp:
deb7b00e877e "planemo upload for repository https://github.com/galaxyproject/tools-iuc/tree/master/tools/trycycler commit 9d7c4277b0f96aacd466f2d497e08edcca3fa238"
iuc
parents:
diff changeset
24 for line in tmp:
deb7b00e877e "planemo upload for repository https://github.com/galaxyproject/tools-iuc/tree/master/tools/trycycler commit 9d7c4277b0f96aacd466f2d497e08edcca3fa238"
iuc
parents:
diff changeset
25 if ">" in line:
deb7b00e877e "planemo upload for repository https://github.com/galaxyproject/tools-iuc/tree/master/tools/trycycler commit 9d7c4277b0f96aacd466f2d497e08edcca3fa238"
iuc
parents:
diff changeset
26 filename = line[1:].strip()
deb7b00e877e "planemo upload for repository https://github.com/galaxyproject/tools-iuc/tree/master/tools/trycycler commit 9d7c4277b0f96aacd466f2d497e08edcca3fa238"
iuc
parents:
diff changeset
27 output_fasta = f"{full_path}{filename}.fasta"
deb7b00e877e "planemo upload for repository https://github.com/galaxyproject/tools-iuc/tree/master/tools/trycycler commit 9d7c4277b0f96aacd466f2d497e08edcca3fa238"
iuc
parents:
diff changeset
28 with open(output_fasta, "a") as handle:
deb7b00e877e "planemo upload for repository https://github.com/galaxyproject/tools-iuc/tree/master/tools/trycycler commit 9d7c4277b0f96aacd466f2d497e08edcca3fa238"
iuc
parents:
diff changeset
29 handle.write(line)
deb7b00e877e "planemo upload for repository https://github.com/galaxyproject/tools-iuc/tree/master/tools/trycycler commit 9d7c4277b0f96aacd466f2d497e08edcca3fa238"
iuc
parents:
diff changeset
30
deb7b00e877e "planemo upload for repository https://github.com/galaxyproject/tools-iuc/tree/master/tools/trycycler commit 9d7c4277b0f96aacd466f2d497e08edcca3fa238"
iuc
parents:
diff changeset
31
deb7b00e877e "planemo upload for repository https://github.com/galaxyproject/tools-iuc/tree/master/tools/trycycler commit 9d7c4277b0f96aacd466f2d497e08edcca3fa238"
iuc
parents:
diff changeset
32 def main():
deb7b00e877e "planemo upload for repository https://github.com/galaxyproject/tools-iuc/tree/master/tools/trycycler commit 9d7c4277b0f96aacd466f2d497e08edcca3fa238"
iuc
parents:
diff changeset
33 if argv[1] == "cluster":
deb7b00e877e "planemo upload for repository https://github.com/galaxyproject/tools-iuc/tree/master/tools/trycycler commit 9d7c4277b0f96aacd466f2d497e08edcca3fa238"
iuc
parents:
diff changeset
34 cluster(argv[2])
deb7b00e877e "planemo upload for repository https://github.com/galaxyproject/tools-iuc/tree/master/tools/trycycler commit 9d7c4277b0f96aacd466f2d497e08edcca3fa238"
iuc
parents:
diff changeset
35 if argv[1] == "reconcile":
deb7b00e877e "planemo upload for repository https://github.com/galaxyproject/tools-iuc/tree/master/tools/trycycler commit 9d7c4277b0f96aacd466f2d497e08edcca3fa238"
iuc
parents:
diff changeset
36 reconcile(argv[2])
deb7b00e877e "planemo upload for repository https://github.com/galaxyproject/tools-iuc/tree/master/tools/trycycler commit 9d7c4277b0f96aacd466f2d497e08edcca3fa238"
iuc
parents:
diff changeset
37
deb7b00e877e "planemo upload for repository https://github.com/galaxyproject/tools-iuc/tree/master/tools/trycycler commit 9d7c4277b0f96aacd466f2d497e08edcca3fa238"
iuc
parents:
diff changeset
38
deb7b00e877e "planemo upload for repository https://github.com/galaxyproject/tools-iuc/tree/master/tools/trycycler commit 9d7c4277b0f96aacd466f2d497e08edcca3fa238"
iuc
parents:
diff changeset
39 if __name__ == "__main__":
deb7b00e877e "planemo upload for repository https://github.com/galaxyproject/tools-iuc/tree/master/tools/trycycler commit 9d7c4277b0f96aacd466f2d497e08edcca3fa238"
iuc
parents:
diff changeset
40 main()