Fix downsample (#222)

* edited file names for downsample fastqs * stringified depth for file name
assemblerflow · Sep 16, 2019 · a2074ca · a2074ca
1 parent 658e6db
commit a2074ca
Show file tree

Hide file tree

Showing 2 changed files with 9 additions and 6 deletions.
diff --git a/flowcraft/generator/templates/downsample_fastq.nf b/flowcraft/generator/templates/downsample_fastq.nf
@@ -17,7 +17,8 @@ process downsample_fastq_{{ pid }} {
     {% include "post.txt" ignore missing %}
 
     tag { "${sample_id}" }
-    publishDir "results/downsample_fastq_{{ pid }}/", pattern: "*_ss.*"
+
+    publishDir "results/downsample_fastq_{{ pid }}/", pattern: "*_ss_*.*"
 
     input:
     set sample_id, file(fastq_pair) from {{ input_channel }}
@@ -27,7 +28,7 @@ process downsample_fastq_{{ pid }} {
     val clear from checkpointClear_{{ pid }}
 
     output:
-    set sample_id, file('*_ss.*') into {{ output_channel }}
+    set sample_id, file('*_ss_*.*') into {{ output_channel }}
     {% with task_name="downsample_fastq" %}
     {%- include "compiler_channels.txt" ignore missing -%}
     {% endwith %}

diff --git a/flowcraft/templates/downsample_fastq.py b/flowcraft/templates/downsample_fastq.py
@@ -134,15 +134,17 @@ def main(sample_id, fastq_pair, genome_size, depth, clear, seed):
         # print ("Writing R1.fq.gz")
         ps = subprocess.Popen(('seqtk', 'sample', parsed_seed, p1, str(ratio)),
                               stdout=subprocess.PIPE)
-        with open('{}_ss.fq.gz'.format(bn1), 'w') as outfile:
+        with open('{}_ss_{}.fq.gz'.format(bn1, str(target_depth)), 'w') as \
+                outfile:
             subprocess.Popen(('gzip', '--fast', '-c'),
                              stdin=ps.stdout, stdout=outfile )
         ps.wait()
 
         # print ("Writing R2.fq.gz")
         ps = subprocess.Popen(('seqtk', 'sample', parsed_seed, p2, str(ratio)),
                               stdout=subprocess.PIPE)
-        with open('{}_ss.fq.gz'.format(bn2), 'w') as outfile:
+        with open('{}_ss_{}.fq.gz'.format(bn2, str(target_depth)), 'w') as \
+                outfile:
             subprocess.Popen(('gzip', '--fast', '-c'),
                              stdin=ps.stdout, stdout=outfile)
         ps.wait()
@@ -157,8 +159,8 @@ def main(sample_id, fastq_pair, genome_size, depth, clear, seed):
                     os.remove(rp)
 
     else:
-        shutil.copy(p1, "{}._ss.fq.gz".format(bn1))
-        shutil.copy(p2, "{}._ss.fq.gz".format(bn2))
+        os.symlink(p1, "{}_ss_{}.fq.gz".format(bn1, str(target_depth)))
+        os.symlink(p2, "{}_ss_{}.fq.gz".format(bn2, str(target_depth)))
 
     # Record the original estimated coverage
     with open(".report.json", "w") as fh: