improve snp text and ploidy bounds

improve snp text
pdimens · Sep 30, 2024 · 33d155c · 33d155c
2 parents 0a9783a + e0590d0
commit 33d155c
Show file tree

Hide file tree

Showing 2 changed files with 5 additions and 8 deletions.
diff --git a/harpy/bin/check_fastq.py b/harpy/bin/check_fastq.py
@@ -33,7 +33,7 @@
 haplotag = re.compile('A[0-9][0-9]C[0-9][0-9]B[0-9][0-9]D[0-9][0-9]')
 bxlast = re.compile('BX:Z:A[0-9][0-9]C[0-9][0-9]B[0-9][0-9]D[0-9][0-9]$')
 
-with pysam.FastxFile(fq_in) as fh:
+with pysam.FastxFile(fq_in, persist=False) as fh:
     N_READS    = 0
     NO_BX       = 0
     BAD_BX      = 0

diff --git a/harpy/snp.py b/harpy/snp.py
@@ -17,12 +17,9 @@ def snp():
     """
     Call SNPs and small indels on alignments
     
-    **Variant Callers**
-    - `mpileup`: call variants using bcftools mpileup
-    - `freebayes`: call variants using freebayes
-
     Provide an additional subcommand `mpileup` or `freebayes` to get more information on using
-    those variant callers. They are both robust variant callers and neither is recommended over the other.
+    those variant callers. They are both robust variant callers, but `freebayes` is recommended when ploidy
+    is greater than **2**.
     """
 
 docstring = {
@@ -52,7 +49,7 @@ def snp():
 @click.option('-x', '--extra-params', type = str, help = 'Additional variant caller parameters, in quotes')
 @click.option('-g', '--genome', type=click.Path(exists=True, dir_okay=False, readable=True), required = True, help = 'Genome assembly for variant calling')
 @click.option('-o', '--output-dir', type = click.Path(exists = False), default = "SNP/mpileup", show_default=True,  help = 'Output directory name')
-@click.option('-n', '--ploidy', default = 2, show_default = True, type=int, help = 'Ploidy of samples')
+@click.option('-n', '--ploidy', default = 2, show_default = True, type=click.IntRange(min = 1, max = 2), help = 'Ploidy of samples')
 @click.option('-p', '--populations', type=click.Path(exists = True, dir_okay=False, readable=True), help = "Tab-delimited file of sample\<tab\>population")
 @click.option('-r', '--regions', type=str, default=50000, show_default=True, help = "Regions where to call variants")
 @click.option('-t', '--threads', default = 4, show_default = True, type = click.IntRange(min = 4, max_open = True), help = 'Number of threads to use')
@@ -156,7 +153,7 @@ def mpileup(inputs, output_dir, regions, genome, threads, populations, ploidy, e
 @click.option('-x', '--extra-params', type = str, help = 'Additional variant caller parameters, in quotes')
 @click.option('-g', '--genome', type=click.Path(exists=True, dir_okay=False, readable=True), required = True, help = 'Genome assembly for variant calling')
 @click.option('-o', '--output-dir', type = click.Path(exists = False), default = "SNP/freebayes", show_default=True,  help = 'Output directory name')
-@click.option('-n', '--ploidy', default = 2, show_default = True, type=int, help = 'Ploidy of samples')
+@click.option('-n', '--ploidy', default = 2, show_default = True, type=click.IntRange(min=1, max_open=True), help = 'Ploidy of samples')
 @click.option('-p', '--populations', type=click.Path(exists = True, dir_okay=False, readable=True), help = "Tab-delimited file of sample\<tab\>population")
 @click.option('-r', '--regions', type=str, default=50000, show_default=True, help = "Regions where to call variants")
 @click.option('-t', '--threads', default = 4, show_default = True, type = click.IntRange(min = 4, max_open = True), help = 'Number of threads to use')