Getting index of structured elements of a list into dictionaries based on regex

Question

I have a list that looks like this:

mylist = ['Probes', 'Gene.symbol', 'Gene.Title', 'GO1', 'GO2', 'GO3', 'ADX_KD_06.ip', 'ADX_KD_24.ip', 'ADX_LG_06.ip', 'ADX_LG_24.ip', 'ADX_LV_06.ip', 'ADX_LV_24.ip', 'ADX_SP_06.ip', 'ADX_SP_24.ip', 'ADX_LN_06.id', 'ALM_LN_06.id', 'ALM_LV_06.ip', 'ALM_SP_06.ip', 'K3SPG_LV_06.ip', 'K3SPG_SP_06.ip', 'KKK_LN_06.id', 'KKK_LV_06.ip', 'KKK_SP_06.ip', 'ENDCN_LV_06.in', 'ENDCN_SP_06.in', 'bCD_LV_06.ip', 'bCD_SP_06.ip', 'ADX_LV_06.id', 'ADX_SP_06.id', 'ALM_LV_06.id', 'ALM_SP_06.id', 'D35_LN_06.id', 'K3SPG_LN_06.id', 'K3_LV_06.id', 'K3_SP_06.id', 'bCD_LN_06.id', 'D35_LV_06.id', 'D35_SP_06.id', 'K3SPG_LV_06.id', 'K3SPG_SP_06.id', 'bCD_LV_06.id', 'bCD_SP_06.id', 'ENDCN_KD_06.in', 'ENDCN_LG_06.in', 'Probes', 'Gene.symbol', 'ADX_KD_06.ip', 'ADX_KD_24.ip', 'ADX_LG_06.ip', 'ADX_LG_24.ip', 'ADX_LV_06.ip', 'ADX_LV_24.ip', 'ADX_SP_06.ip', 'ADX_SP_24.ip', 'ADX_LN_06.id', 'ALM_LN_06.id', 'ALM_LV_06.ip', 'ALM_SP_06.ip', 'K3SPG_LV_06.ip', 'K3SPG_SP_06.ip', 'KKK_LN_06.id', 'KKK_LV_06.ip', 'KKK_SP_06.ip', 'ENDCN_LV_06.in', 'ENDCN_SP_06.in', 'bCD_LV_06.ip', 'bCD_SP_06.ip', 'ADX_LV_06.id', 'ADX_SP_06.id', 'ALM_LV_06.id', 'ALM_SP_06.id', 'D35_LN_06.id', 'K3SPG_LN_06.id', 'K3_LV_06.id', 'K3_SP_06.id', 'bCD_LN_06.id', 'D35_LV_06.id', 'D35_SP_06.id', 'K3SPG_LV_06.id', 'K3SPG_SP_06.id', 'bCD_LV_06.id', 'bCD_SP_06.id', 'ENDCN_KD_06.in', 'ENDCN_LG_06.in']
# index =    0,        1,             2,          3,      4,    5,      6, ....

We can see the list above to be separated into two parts, delimited by 'Probes', 'Gene.symbol', 'Gene.Title', 'GO1', 'GO2', 'GO3' and 'Probes', 'Gene.symbol'.

The regex for the components of the first and second part is:

([\w\d]+)_(\w\w)_(\d\d)\.(\w\w)
  rg1      rg2     rg3    rg4

Which should match string like this ADX_SP_06.ip or K3SPG_LN_06.id

What I want to do is to create a dictionary with keys formed by rg1 + rg3 where the values stored the index of the members in first and second part of mylist. The results will look like this.

first_part_dict = {
      'bCD06'    : { 25, 26, 35, 40, 41 },
      'ADX06'    : { 6,  8,  10, 12, 14, 27, 28 },
      'ADX24'    : { 7, 9, 11, 13 },
      'ALM06'    : { 15, 16, 17, 29, 30 },
      'K3SPG06'  : { 18, 19, 32, 38, 39 },
      'KKK06'    : { 20, 21, 22 },
      'K306'     : { 33, 34},
      'ENDCN06'  : { 23, 24, 42,43 },
      'D3506'    : { 31, 36, 37 },
}

second_part_dict = {
      'bCD06'    : { 65, 66, 75, 80, 81 },
      'ADX06'    : { 46, 48, 50, 52, 54, 67, 68},
      'ADX24'    : { 47, 49, 51, 53 },
      'ALM06'    : { 55, 56, 57, 69, 70 },
      'K3SPG06'  : { 58, 59, 72, 78, 79 },
      'KKK06'    : { 60, 61, 62 },
      'K306'     : { 73, 74},
      'ENDCN06'  : { 63, 64, 82, 83},
      'D3506'    : { 71, 76, 77 },
}

How can I achieve that with Python?

Answer 1

This isn't very elegant, but it should get the job done. I have assumed based on your expected output that you meant for the key to be rg1 + rg3, not rg1 + rg2.

import re, collections

mylist = ['Probes', 'Gene.symbol', 'Gene.Title', 'GO1', 'GO2', 'GO3', 'ADX_KD_06.ip', 'ADX_KD_24.ip', 'ADX_LG_06.ip', 'ADX_LG_24.ip', 'ADX_LV_06.ip', 'ADX_LV_24.ip', 'ADX_SP_06.ip', 'ADX_SP_24.ip', 'ADX_LN_06.id', 'ALM_LN_06.id', 'ALM_LV_06.ip', 'ALM_SP_06.ip', 'K3SPG_LV_06.ip', 'K3SPG_SP_06.ip', 'KKK_LN_06.id', 'KKK_LV_06.ip', 'KKK_SP_06.ip', 'ENDCN_LV_06.in', 'ENDCN_SP_06.in', 'bCD_LV_06.ip', 'bCD_SP_06.ip', 'ADX_LV_06.id', 'ADX_SP_06.id', 'ALM_LV_06.id', 'ALM_SP_06.id', 'D35_LN_06.id', 'K3SPG_LN_06.id', 'K3_LV_06.id', 'K3_SP_06.id', 'bCD_LN_06.id', 'D35_LV_06.id', 'D35_SP_06.id', 'K3SPG_LV_06.id', 'K3SPG_SP_06.id', 'bCD_LV_06.id', 'bCD_SP_06.id', 'ENDCN_KD_06.in', 'ENDCN_LG_06.in', 'Probes', 'Gene.symbol', 'ADX_KD_06.ip', 'ADX_KD_24.ip', 'ADX_LG_06.ip', 'ADX_LG_24.ip', 'ADX_LV_06.ip', 'ADX_LV_24.ip', 'ADX_SP_06.ip', 'ADX_SP_24.ip', 'ADX_LN_06.id', 'ALM_LN_06.id', 'ALM_LV_06.ip', 'ALM_SP_06.ip', 'K3SPG_LV_06.ip', 'K3SPG_SP_06.ip', 'KKK_LN_06.id', 'KKK_LV_06.ip', 'KKK_SP_06.ip', 'ENDCN_LV_06.in', 'ENDCN_SP_06.in', 'bCD_LV_06.ip', 'bCD_SP_06.ip', 'ADX_LV_06.id', 'ADX_SP_06.id', 'ALM_LV_06.id', 'ALM_SP_06.id', 'D35_LN_06.id', 'K3SPG_LN_06.id', 'K3_LV_06.id', 'K3_SP_06.id', 'bCD_LN_06.id', 'D35_LV_06.id', 'D35_SP_06.id', 'K3SPG_LV_06.id', 'K3SPG_SP_06.id', 'bCD_LV_06.id', 'bCD_SP_06.id', 'ENDCN_KD_06.in', 'ENDCN_LG_06.in']

regex = re.compile(r'([\w\d]+)_(\w\w)_(\d\d)\.(\w\w)')
first_part_dict = collections.defaultdict(list)
second_part_dict = collections.defaultdict(list)
# second instance of 'Probes', to separate the first and second parts
cutoff_index = mylist.index('Probes', 1) 

for i, string in enumerate(mylist):
    matched = regex.match(string)
    if not matched:
        continue
    rg1, rg2, rg3, rg4 = matched.groups()
    key = rg1 + rg3
    if i < cutoff_index:
        first_part_dict[key].append(i)
    else:
        second_part_dict[key].append(i)

Result:

>>> first_part_dict
defaultdict(<class 'list'>, {'ALM06': [15, 16, 17, 29, 30], 'K3SPG06': [18, 19, 32, 38, 39], 'bCD06': [25, 26, 35, 40, 41], 'ADX24': [7, 9, 11, 13], 'ENDCN06': [23, 24, 42, 43], 'KKK06': [20, 21, 22], 'K306': [33, 34], 'ADX06': [6, 8, 10, 12, 14, 27, 28], 'D3506': [31, 36, 37]})
>>> second_part_dict
defaultdict(<class 'list'>, {'ALM06': [55, 56, 57, 69, 70], 'K3SPG06': [58, 59, 72, 78, 79], 'bCD06': [65, 66, 75, 80, 81], 'ADX24': [47, 49, 51, 53], 'ENDCN06': [63, 64, 82, 83], 'KKK06': [60, 61, 62], 'K306': [73, 74], 'ADX06': [46, 48, 50, 52, 54, 67, 68], 'D3506': [71, 76, 77]})