CytoTargetDB/basic_functions.py at main · MorrissyLab/CytoTargetDB · 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
from object_classes import ClinicalTrial, ArmGroup, Intervention
from typing import List, Tuple, Optional
import pandas as pd
from thefuzz import fuzz

def parse_clinical_trial(data, trial_source: Optional[str] = None, annotation_status: Optional[str] = None) -> ClinicalTrial:
    """Parse clinical trial data"""

    protocol = data['protocolSection']
    derived = data.get('derivedSection', {})

    # Get mesh terms from the meshes list
    conditions_mesh_terms = []
    for mesh in derived.get('conditionBrowseModule', {}).get('meshes', []):
        if 'term' in mesh:
            conditions_mesh_terms.append(mesh['term'])

    interventions_mesh_terms = []
    for mesh in derived.get('interventionBrowseModule', {}).get('meshes', []):
        if 'term' in mesh:
            interventions_mesh_terms.append(mesh['term'])

    # Extract interventions
    interventions = []
    for inv in protocol.get('armsInterventionsModule', {}).get('interventions', []):
        intervention: Intervention = {
            "type": inv.get('type', ''),
            "name": inv.get('name', ''),
            "description": inv.get('description', ''),
            "arm_group_labels": inv.get('armGroupLabels', []),
            "other_names": inv.get('otherNames', [])
        }
        interventions.append(intervention)

    # Extract arm groups
    arm_groups = []
    for arm in protocol.get('armsInterventionsModule', {}).get('armGroups', []):
        arm_group: ArmGroup = {
            "label": arm.get('label', ''),
            "type": arm.get('type', ''),
            "description": arm.get('description', ''),
            "intervention_names": arm.get('interventionNames', [])
        }
        arm_groups.append(arm_group)

    # Create main trial object
    trial: ClinicalTrial = {
        "nct_id": protocol['identificationModule']['nctId'],
        "brief_title": protocol['identificationModule']['briefTitle'],
        "official_title": protocol['identificationModule']['officialTitle'],

        "brief_summary": protocol.get('descriptionModule', {}).get('briefSummary', ''),
        "detailed_description": protocol.get('descriptionModule', {}).get('detailedDescription', ''),
        "conditions": protocol.get('conditionsModule', {}).get('conditions', []),
        "keywords": protocol.get('conditionsModule', {}).get('keywords', []),

        "study_type": protocol.get('designModule', {}).get('studyType', ''),

        "arm_groups": arm_groups,
        "interventions": interventions,

        "condition_mesh_terms": conditions_mesh_terms,
        "intervention_mesh_terms": interventions_mesh_terms,

        "eligibility_criteria": protocol.get('eligibilityModule', {}).get('eligibilityCriteria', ''),

        "trial_source": trial_source,
        "annotation_status": annotation_status,

    }
    return trial


def match_drugs_to_nci_thesaurus(drug, nci_df) -> dict:
    """
    Match drug names against NCI thesaurus and return matches with codes and definitions
    """

    best_match = None
    best_score = 0

    # Search through display names and synonyms
    for _, row in nci_df.iterrows():
        # Check display name
        score = fuzz.ratio(drug.lower(), str(row['display name']).lower())
        if score > best_score:
            best_score = score
            best_match = row

        # Check synonyms if present
        if pd.notna(row['synonyms']):
            for synonym in str(row['synonyms']).split('|'):
                score = fuzz.ratio(drug.lower(), synonym.lower())
                if score > best_score:
                    best_score = score
                    best_match = row
    match = {}
    if best_score > 97:  # Only include true matches
        match = {
            'drug_name': drug,
            'nci_code': best_match['code'],
            'nci_term': best_match['display name'],
            'definition': best_match['definition']
        }

    return match