Mercurial > repos > davidvanzessen > shm_csr

comparison mutation_column_checker.py @ 78:aff3ba86ef7a draft

Find changesets by keywords (author, files, the commit message), revision number or hash, or revset expression.
Uploaded
author davidvanzessen
date Mon, 31 Aug 2020 11:20:08 -0400
parents
children
comparison
equal deleted inserted replaced
77:58d2377b507d 78:aff3ba86ef7a
1 import re
2
3 mutationMatcher = re.compile("^([nactg])(\d+).([nactg]),?[ ]?([A-Z])?(\d+)?[>]?([A-Z;])?(.*)?")
4
5 with open("7_V-REGION-mutation-and-AA-change-table.txt", 'r') as file_handle:
6 first = True
7 fr3_index = -1
8 for i, line in enumerate(file_handle):
9 line_split = line.split("\t")
10 if first:
11 fr3_index = line_split.index("FR3-IMGT")
12 first = False
13 continue
14
15 if len(line_split) < fr3_index:
16 continue
17
18 fr3_data = line_split[fr3_index]
19 if len(fr3_data) > 5:
20 try:
21 test = [mutationMatcher.match(x).groups() for x in fr3_data.split("|") if x]
22 except:
23 print(line_split[1])
24 print("Something went wrong at line {line} with:".format(line=line_split[0]))
25 #print([x for x in fr3_data.split("|") if not mutationMatcher.match(x)])
26 if i % 100000 == 0:
27 print(i)