biomappings/scripts/generate_cl_mesh_mappings.py at main · biopragmatics/biomappings · 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
"""Generate mappings from CL."""

import re

import gilda
import obonet
from bioregistry import NormalizedNamableReference
from curies.vocabulary import exact_match, lexical_matching_process
from indra.databases import mesh_client
from sssom_pydantic import MappingTool, SemanticMapping

from biomappings.resources import append_predictions
from biomappings.utils import get_script_url

g = obonet.read_obo(
    "https://raw.githubusercontent.com/obophenotype/cell-ontology/master/cl-basic.obo"
)

mesh_tree_pattern = re.compile(r"MESH:[A-Z][0-9]+\.[0-9.]+")
mesh_id_pattern = re.compile(r"MESH:[CD][0-9]+")

mappings = {}
for node, data in g.nodes(data=True):
    if not node.startswith("CL:"):
        continue

    has_mesh_id = False
    for value in [data.get("def", ""), *data.get("synonym", []), *data.get("xref", [])]:
        if re.findall(mesh_tree_pattern, value) or re.findall(mesh_id_pattern, value):
            has_mesh_id = True
            break

    if has_mesh_id:
        continue

    matches = gilda.ground(data["name"])  # type:ignore[no-untyped-call]
    if not matches:
        if data["name"].endswith(" cells"):
            matches = gilda.ground(data["name"].replace(" cells", ""))  # type:ignore[no-untyped-call]
        elif data["name"].endswith(" cell"):
            matches = gilda.ground(data["name"].replace(" cell", ""))  # type:ignore[no-untyped-call]
    if not matches:
        continue

    mesh_ids = set()
    for match in matches:
        groundings = match.get_groundings()
        mesh_ids |= {id for ns, id in groundings if ns == "MESH"}
    if len(mesh_ids) > 1:
        print(f"Multiple MESH IDs for {node}")
    elif len(mesh_ids) == 1:
        mesh_id = next(iter(mesh_ids))
        mappings[node] = mesh_id

print(f"Found {len(mappings)} CL->MESH mappings.")

provenance = get_script_url(__file__)

predictions = []
for cl_id, mesh_id in mappings.items():
    pred = SemanticMapping(
        subject=NormalizedNamableReference(
            prefix="cl",
            identifier=cl_id,
            name=g.nodes[cl_id]["name"],
        ),
        predicate=exact_match,
        object=NormalizedNamableReference(
            prefix="mesh",
            identifier=mesh_id,
            name=mesh_client.get_mesh_name(mesh_id),
        ),
        justification=lexical_matching_process,
        confidence=0.9,
        mapping_tool=MappingTool(name=provenance),
    )
    predictions.append(pred)

append_predictions(predictions)