Skip to content
Snippets Groups Projects
Select Git revision
  • main default protected
1 result
Blame History Permalink
msms_processing.py 3.41 KiB
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98 






import pandas as pd
import numpy as np
import random
from constant import ALPHABET_UNMOD, ALPHABET_UNMOD_REV

def load_data(msms_filet_path='data/msms_plasma.txt', score_treshold=70):
    data = pd.read_csv(msms_filet_path, sep='\t')
    data_compact = data[['Modified sequence', 'Length', 'Charge', 'Retention time', 'Score', 'Matches', 'Intensities']]
    data_filtered = data_compact[data_compact['Score'] > score_treshold]
    data_filtered = data_filtered[data_filtered['Length'] < 26]
    data_filtered = data_filtered[['Modified sequence', 'Retention time', 'Score']]
    data_filtered['Modified sequence']=data_filtered['Modified sequence'].map(lambda x : x[1:-1])
    data_filtered['Modified sequence']=data_filtered['Modified sequence'].map(convert_mod_to_prosit)
    data_filtered.rename(columns={'Modified sequence': 'sequence', 'Retention time': 'irt_scaled', 'Score':'score'}, inplace=True)
    return data_filtered


def convert(l):
    return [num_int for num_str in l.split() for num_int in (
        lambda x: [float(x.replace('[', '').replace(']', ''))] if x.replace('.', '').replace('[', '').replace(']',
                                                                                                              '').replace(
            'e+', '').isdigit() else [])(
        num_str)]

def convert_mod_to_prosit(s):
    return s.replace('M(ox)','-OxM-').replace('(ac)','@')

def numerical_to_alphabetical(arr):
    seq = ''
    for i in range(len(arr)):
        seq+=ALPHABET_UNMOD_REV[arr[i]]
    return seq

def filter_acetyl(df):
    df2 = df.copy()
    df2['acetyl']=df['sequence'].map(lambda x: '@' in x)
    df2 = df2[df2['acetyl']==False]
    df_final = df2.drop('acetyl', axis=1)
    df_final.reset_index(drop=True, inplace=True)
    return df_final

def filter_cysteine(df):
    df2 = df.copy()
    df2['cysteine']=df['sequence'].map(lambda x: 'C' in x)
    df2 = df2[df2['cysteine']==False]
    df_final = df2.drop('cysteine', axis=1)
    df_final.reset_index(drop=True, inplace=True)
    return df_final

def add_split_column(data, split=(0.7,0.15,0.15)):
    num_data = data.shape[0]
    train_num = np.floor(num_data*split[0])
    val_num = np.floor(num_data*(split[0]+split[1]))
    num_data=0
    list_train=[]
    list_test=[]
    list_val=[]
    groups = [df for _, df in data.groupby('sequence')]
    random.shuffle(groups)

    data = pd.concat(groups).reset_index(drop=True)

    groups = data.groupby('sequence',sort=False)
    for seq, gr in groups:

        num_data+= gr.shape[0]

        if num_data<train_num:
            list_train.append(gr)
        elif num_data<val_num:
            list_val.append(gr)
        else:
            list_test.append(gr)



    dataset_train = pd.concat(list_train, ignore_index=True)
    dataset_train['state'] = 'train'
    dataset_val = pd.concat(list_val, ignore_index=True)
    dataset_val['state'] = 'validation'
    dataset_test = pd.concat(list_test, ignore_index=True)
    dataset_test['state'] = 'holdout'

    data_split = pd.concat([dataset_train, dataset_val, dataset_test], ignore_index=True)
    data_split.reset_index(drop=True, inplace=True)

    return data_split

def main():
    df_plasma = load_data('data_PXD006109/msms_plasma.txt', 70)
    merged_df = pd.concat([df_plasma], ignore_index=True)
    final_df = add_split_column(merged_df)
    final_df = filter_acetyl(final_df)
    final_df.to_csv('data_PXD006109/data_plasma.csv', index=False)


if __name__ == '__main__':
    main()