compile-protein-interactions/ner_relation.py at main · hackbio-ca/compile-protein-interactions · GitHub

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
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
from indra.sources import trips
import pandas as pd
import json
import re
import spacy
import scispacy
import requests
import xml.etree.ElementTree as ET
import json

def find_key_sentences(abstract, ppi_terms):

    if "protein" in abstract:

        # Find sentences containing any PPI keyword
        sentences = re.split(r'(?<=[.!?])\s+', abstract)
        matched_sentences = [
            sentence for sentence in sentences
            if re.search(r"\b(" + "|".join(ppi_terms) + r")\b", sentence, re.IGNORECASE)
        ]

        return matched_sentences

# Read abstract data
with open('pmid_to_abstract_updated.json', 'r') as f:
    data = json.load(f)

# Load model
nlp = spacy.load("en_ner_jnlpba_md")


ids=["40973402",
     "40973401",
     "40966293",
     "40915373",
     "40587559",
     "40913132",
     "40758965",
     "40752663",
     "40742404",
     "40721030",
     "40713730",
     "40707822",
     "40448625",
     "40358171",
     "40347153",
     "40340056",
     "40296620",
     "39695808",
     "39680531", #
     "40676669",
     "40927314",
     "40551140",
     "40818508",
     "40815569",
     "40441521",
     "40373772",
     "40362335",
     "40247363",
     "40223243",
     "40212965",
     "40205047",
     "40183841",
     "40176187",
     "40175501",
     "40131356",
     "40301248",
     "40318722",
     "40337095",
     "40337551",
     "40081988",
     "39759399",
     "39775030",
     "39870805",
     "39902677",
     "39952508",
     "39975251",
     "40001482",
     "40041897",
     "40368227"]

cells_pmid = []

f = open('data_new.csv','w')

for id in ids:

    # Run Relation Extraction
    ppi_terms = ["binds", "associates with", "interacts with", "activates", "inhibits", "interact with", "strong binding"]
    key_sentences = find_key_sentences(data[id], ppi_terms)

    if not key_sentences:
        print("Couldn't find any key sentences")
        continue
    else:
        print(key_sentences)

    for s in key_sentences:
        tp = trips.process_text(s)

        # Find first protein interaction
        if tp.statements:

            interactions = [str(i) for i in tp.statements]
            ppi_sentence = s
            print(tp.statements)
            break

        else:
            interactions=[]


    if not interactions:
        # Didn't find any protein-protein interactions
        print("Couldn't find any ppis")
        continue

    # Get protein a and protein b of each interaction
    for i in interactions:
        a_start = i.find('(')
        a_end = i.find('(', a_start + 1)

        b_start = i.find(',')
        b_end = i.find('(', b_start+1)

        protein_a = i[a_start+1:a_end]
        protein_b = i[b_start+2:b_end]

        protein_a = protein_a.lower()
        protein_b = protein_b.lower()

        print(f'A: {protein_a} B:{protein_b}')

    # Run NER
    abstract = data[id]
    doc = nlp(abstract)
    entity_types = ['CELL_LINE', 'CELL_TYPE']
    unique_cells = set() # unique cell names

    for ent in doc.ents:
        if ent.label_ in entity_types:

            cell_name = ent.text

            # Make lowercase and replace space with underscore
            cell_name_clean = cell_name.lower().replace(" ", "_")

            # If you found a new cell for ppi, add; otherwise, discard
            if cell_name_clean not in unique_cells:

                unique_cells.add(cell_name_clean)
                hypernode = f'{protein_a}-{protein_b}-{id}'
                f.write(f'"{protein_a}","{protein_b}","{ppi_sentence}","{cell_name_clean}","{id}","{hypernode}"\n')


f.close()