Merge pull request #32 from EBIvariation/remove_non_ATGC

Remove variant that have non ATGC bases in the reference
EBIvariation · Aug 23, 2021 · f44fac1 · f44fac1
2 parents f7303a5 + bc3661f
commit f44fac1
Show file tree

Hide file tree

Showing 2 changed files with 9 additions and 7 deletions.
diff --git a/conda.yml b/conda.yml
@@ -4,10 +4,8 @@ channels:
   - conda-forge
   - bioconda
 dependencies:
-  - bedtools=2.29.2
-  - bowtie2=2.4.1
-  - minimap2=2.17
-  - samtools=1.9
-  - bcftools=1.9
-  - bedops=2.4.39
-  - tabix=0.2.6
+  - bedtools
+  - minimap2
+  - samtools
+  - bcftools
+  - tabix
diff --git a/variant_remapping_tools/reads_to_remapped_variants.py b/variant_remapping_tools/reads_to_remapped_variants.py
@@ -8,6 +8,7 @@
 from Bio.Alphabet import generic_dna
 import pysam
 
+nucleotide_alphabet = {'A', 'T', 'C', 'G'}
 
 def reverse_complement(sequence):
     return str(Seq(sequence, generic_dna).reverse_complement())
@@ -27,6 +28,9 @@ def calculate_new_variant_definition(left_read, right_read, ref_fasta, original_
     new_ref = fetch_bases(ref_fasta, left_read.reference_name, left_read.reference_end + 1,
                           right_read.reference_start - left_read.reference_end).upper()
 
+    if len(set(new_ref).difference(nucleotide_alphabet)) != 0 :
+        failure_reason = 'Reference Allele not in ACGT'
+
     new_pos = left_read.reference_end + 1
 
     # 1. Handle reference strand change