tanu/LSHTM_analysis
1
0
Fork 0
Code Issues Pull requests Releases Wiki Activity

adding clean files for rerrun 35k dataset

Browse source
This commit is contained in:
Tanushree Tunstall 2020-07-07 18:28:55 +01:00
parent 0973717287
commit 8f460347b4
32 changed files with 157 additions and 44550 deletions
Download patch file Download diff file Expand all files Collapse all files

99
scripts/find_missense.py
View file

@ -1,99 +0,0 @@
#!/usr/bin/env python3
import pandas as pd
DEBUG = False
#%%
#def find_missense(test_df, ref_allele1, alt_allele0):
def find_missense(test_df, ref_allele_column, alt_allele_column, n_diff_colname = 'n_diff', tot_diff_colname = 'tot_diff', ref_a_colname = 'ref_allele', alt_a_colname = 'alt_allele'):
"""Find mismatches in pairwise comparison of strings b/w col_a and col_b
Case insensitive, converts strings to uppercase before comparison
@test_df: df containing columns to compare
@type: pandas df
@ref_allele_column: column containing ref allele str
@type: str (converts to uppercase)
@alt_allele_column: column containing alt_allele str
@type: str (converts to uppercase)
@n_diff_colname: user defined colname for no. of char diff b/w ref_allele_str and alt_allele_str
@type: str
@tot_diff_colname: user defined colname abs diff to indicate if strings are of equal length
@type: str
@ref_a_colname: user defined colname containing extracted referece allele
@type: str
@alt_a_colname: user defined colname containing extracted alt allele
@type: str
returns df: with 4 columns. If column names clash, the function column
name will override original column
@rtype: pandas df
"""
for ind, val in test_df.iterrows():
if DEBUG:
print('index:', ind, 'value:', val
, '\n============================================================')
ref_a = val[ref_allele_column].upper()
alt_a = val[alt_allele_column].upper()
if DEBUG:
print('ref_allele_string:', ref_a, 'alt_allele_string:', alt_a)
difference = sum(1 for e in zip(ref_a, alt_a) if e[0] != e[1])
test_df.at[ind, n_diff_colname] = difference # adding column
tot_difference = difference + abs(len(ref_a) - len(alt_a))
test_df.at[ind, tot_diff_colname] = tot_difference # adding column
if difference != tot_difference:
print('WARNING: lengths of ref_allele and alt_allele differ at index:', ind
, '\nNon-missense muts detected')
# Now finding the mismatched char
ref_aln = ''
alt_aln = ''
if ref_a == alt_a:
##test_df.at[ind, 'ref_allele'] = 'no_change' # adding column
##test_df.at[ind, 'alt_allele'] = 'no_change' # adding column
test_df.at[ind, ref_a_colname] = 'no_change' # adding column
test_df.at[ind, alt_a_colname] = 'no_change' # adding column
elif len(ref_a) == len(alt_a) and len(ref_a) > 0:
print('ref:', ref_a, 'alt:', alt_a)
for n in range(len(ref_a)):
if ref_a[n] != alt_a[n]:
ref_aln += ref_a[n]
alt_aln += alt_a[n]
##test_df.at[ind, 'ref_allele'] = ref_aln
##test_df.at[ind, 'alt_allele'] = alt_aln
test_df.at[ind, ref_a_colname] = ref_aln
test_df.at[ind, alt_a_colname] = alt_aln
print('ref:', ref_aln)
print('alt:', alt_aln)
else:
##test_df.at[ind, 'ref_allele'] = 'ERROR_Not_nsSNP'
##test_df.at[ind, 'alt_allele'] = 'ERROR_Not_nsSNP'
test_df.at[ind, ref_a_colname] = 'ERROR_Not_nsSNP'
test_df.at[ind, alt_a_colname] = 'ERROR_Not_nsSNP'
return test_df
#========================================
# a representative example
eg_df = {'chromosome_number': [2288719, 2288766, 2288775, 2288779, 2288827, 1111111, 2222222],
'ref_allele1': ['Tc', 'AG', 'AGCACCCTG', 'CCCTGGTGGCC', 'CACA', 'AA', 'CAT'],
'alt_allele0': ['CC', 'CA', 'GGCACCCTGZ','TCCTGGTGGCCAAD', 'TACA', 'AA', 'TCZ']}
# snippet of actual data
#eg_df = pd.read_csv('pnca_assoc.txt', sep = '\t', nrows = 10, header = 0, index_col = False)
eg_df = pd.DataFrame(eg_df)
def main():
#find_missense(eg_df, ref_allele1 = 'ref_allele', alt_allele0 = 'alt_allele')
find_missense(test_df = eg_df, ref_allele_column = 'ref_allele1', alt_allele_column = 'alt_allele0')
print(eg_df)
if __name__ == '__main__':
main()