def setUpClass(self):
''' __init__ method for class obj '''
self.data_path = os.path.abspath(__file__ + '/../' + \
'data/test_find_peptide_variants/')
annotation_path = self.data_path + '/gencode_min.gtf'
genomefa_path = self.data_path + '/GRCh38_limited_chr7.fa.gz'
cosmicdb_path = self.data_path + \
'/CosmicGenomeScreensMutantExport.min.tsv'
self.input_path = self.data_path + '/vcf/'
self.input_paths = [
self.input_path + x for x in os.listdir(self.input_path)
]
self.cosmic_df = pd.read_csv(cosmicdb_path, sep='\t')
self.annotation_df = pd.read_csv(annotation_path, sep='\t', skiprows=5)
self.genome_faidx = Fasta(genomefa_path)
self.annotation_genome_tree = GenomeIntervalTree(
lambda feat: feat.pos,
(GFFFeature(row) for _, row in self.annotation_df.iterrows()))
self.protein_variant_predictor = ProteinVariantPredictor(
self.annotation_genome_tree, self.genome_faidx)
self.aa_mutation_finder = AminoAcidMutationFinder(self.cosmic_df,
self.annotation_df,
self.genome_faidx,
cov_bool=1)
def setUpClass(self):
''' __init__ method for FindAAMutationsTester class '''
self.data_path = os.path.abspath(__file__ + \
'/../' + 'data/test_find_peptide_variants/')
self.cosmicdb_path = self.data_path + '/cosmic_min.tsv'
self.annotation_path = self.data_path + '/hg38-plus.min.gtf'
self.genomefa_path = self.data_path + '/GRCh38_limited_chr7.fa.gz'
self.cov_bool = 1
self.num_processes = 2 # this is wierd
self.outpath = self.data_path + '/test_out.csv'
self.input_path = self.data_path + '/vcf/'
self.input_paths = [
self.input_path + x for x in os.listdir(self.input_path)
]
cosmic_df = None
annotation_df = pd.read_csv(self.annotation_path, sep='\t', skiprows=5)
genome_faidx = Fasta(self.genomefa_path)
self.aa_mutation_finder = AminoAcidMutationFinder(
cosmic_df, annotation_df, genome_faidx, self.cov_bool)
def test_extract_coverage(self):
''' testing the extract_coverage() subroutine within
find_peptide_variants.find_cell_gene_aa_mutations() '''
def init_process(curr_aa_mutation_finder):
global current_process_aa_mutation_finder
current_process_aa_mutation_finder = curr_aa_mutation_finder
def process_cell(path):
my_obj = current_process_aa_mutation_finder. \
find_cell_gene_aa_mutations(path=path)
return (my_obj)
# need to rebuild this with larger version of COSMIC
cosmicdb_path = self.data_path + '/cosmic_min.tsv'
cosmic_df = pd.read_csv(cosmicdb_path, sep='\t')
curr_aa_mutation_finder = AminoAcidMutationFinder(cosmic_df,
self.annotation_df,
self.genome_faidx,
cov_bool=1)
for vcf_path in self.input_paths:
curr_vcf = vcf_path.strip(self.input_path)
init_process(curr_aa_mutation_finder)
gene_aa_mutations = process_cell(vcf_path)
if 'A1' in curr_vcf:
k = list(gene_aa_mutations)
assert 'EGFR' in k
# take a look at the variant coverage string
# this is horrible -- I know
v = gene_aa_mutations.get('EGFR')
vl = list(v)
ve = vl[0]
vc = ve.split(',')[1]
assert vc == '[2:0]'
elif 'A2' in curr_vcf:
k = list(gene_aa_mutations)
v = gene_aa_mutations.get('EGFR')
vl = list(v)
ve = vl[0]
vc = ve.split(',')[1]
assert vc == '[0:2]' or vc == '[3:2:1]'
else:
k = list(gene_aa_mutations)
assert not k
def test_basic_cmp():
''' does find_all_mutations return w/o error, redux
this one has an expected vs. actual file compare step
pytest keeps telling me filecmp.cmp() is outdated? '''
from cerebra.find_peptide_variants import AminoAcidMutationFinder
data_path = os.path.abspath(__file__ + '/../' +
'data/test_find_peptide_variants/')
genomefa_path = data_path + '/GRCh38_limited_chr7.fa'
annotation_path = data_path + '/gencode_min.gtf'
coverage = 0
num_processes = 1 # want to include that multiprocessing module
outpath = data_path + '/test_out.csv'
input_path = data_path + '/vcf/'
input_paths = [input_path + x for x in os.listdir(input_path)]
cosmic_df = None
annotation_df = pd.read_csv(annotation_path, sep='\t', skiprows=5)
genome_faidx = Fasta(genomefa_path)
aa_mutation_finder = AminoAcidMutationFinder(cosmic_df, annotation_df,
genome_faidx, coverage)
results_df = aa_mutation_finder.find_transcript_mutations(
paths=input_paths, processes=num_processes)
results_df = results_df.sort_index() # row sort
results_df = results_df.sort_index(axis=1) # column sort
results_df.to_csv(outpath)
assert os.path.isfile(outpath)
outdf = pd.read_csv(outpath, index_col=0)
expect_index = ['A1', 'A2', 'A3', 'A4', 'A5']
expect_cols = ['EGFR']
assert list(outdf.index) == expect_index
assert list(outdf.columns) == expect_cols
indel_ensps = [
'ENSP00000275493.2:p.(Leu858delinsArgTrp)',
'ENSP00000395243.3:p.(Leu813delinsArgTrp)',
'ENSP00000415559.1:p.(Leu813delinsArgTrp)'
]
snp_ensps = [
'ENSP00000415559.1:p.(Leu813Arg)',
'ENSP00000395243.3:p.(Leu813Arg)',
'ENSP00000275493.2:p.(Leu858Arg)',
]
a1_matches = outdf.loc['A1']['EGFR']
a2_matches = outdf.loc['A2']['EGFR']
# this is a bit backwards but ok considering the wierd format of outfile
for x in indel_ensps:
assert x in a1_matches
for y in snp_ensps:
assert y in a1_matches
for z in snp_ensps:
assert z in a2_matches
#teardown
os.remove(data_path + "/test_out.csv")