Spaces:
Sleeping
Sleeping
File size: 9,960 Bytes
33d7314 | 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 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 | from __future__ import annotations
from pathlib import Path
from typing import Any
import pandas as pd
from .adapter import prepare_from_archive, prepare_from_source
from .common import profile_for_pubtime
COMPARATOR_FIELDS = (
"phase",
"primary_purpose",
"intervention_model",
"allocation",
"masking",
"gender",
"number_of_arms",
"has_dmc",
"has_us_facility",
)
def summarize_from_archive(profile: dict[str, Any], project_root: Path) -> dict[str, Any]:
prepared, source_label, source_type = prepare_from_archive(profile.get("domain", ""), project_root)
return _summarize_prepared(profile, prepared, source_label, source_type)
def summarize_from_zip(profile: dict[str, Any], zip_path: Path) -> dict[str, Any]:
prepared, source_label, source_type = prepare_from_source(profile.get("domain", ""), zip_path, zip_path.name)
return _summarize_prepared(profile, prepared, source_label, source_type)
def _summarize_prepared(
profile: dict[str, Any],
prepared,
source_label: str,
source_type: str,
) -> dict[str, Any]:
rows = _frame_to_rows(prepared.survival_df)
pubtime_profile = profile_for_pubtime(profile)
matched = _matched_rows(pubtime_profile, rows)
comparator_rows = matched if len(matched) >= 20 else rows
return {
"domain": prepared.domain,
"source_archive": source_label,
"source_type": source_type,
"raw_rows": prepared.raw_rows,
"analytic_rows": prepared.analytic_rows,
"domain_rows": prepared.survival_rows,
"matched_rows": len(matched),
"used_rows": len(comparator_rows),
"match_strategy": _match_strategy(len(matched)),
"pubtime_preparation": {
"predictors": list(prepared.predictors),
"high_missing_predictors": list(prepared.high_missing_predictors),
"cox_model": prepared.cox_model_status,
},
"summary": _summarize_rows(comparator_rows),
"comparison": _compare_profile(profile, comparator_rows),
"examples": _example_rows(pubtime_profile, comparator_rows),
}
def _frame_to_rows(frame: pd.DataFrame) -> tuple[dict[str, Any], ...]:
if frame.empty:
return tuple()
wanted = {
"nct_id",
"brief_title",
"phase",
"primary_purpose",
"intervention_model",
"allocation",
"masking",
"enrollment",
"actual_duration",
"number_of_facilities",
"number_of_arms",
"number_of_primary_outcomes_to_measure",
"number_of_secondary_outcomes_to_measure",
"has_dmc",
"has_us_facility",
"were_results_reported",
"pub_date",
"event_pub",
"time_to_pub",
"result_count",
}
available = [column for column in wanted if column in frame.columns]
rows = frame[available].where(pd.notna(frame[available]), "").to_dict("records")
return tuple(rows)
def _matched_rows(profile: dict[str, Any], rows: tuple[dict[str, Any], ...]) -> list[dict[str, Any]]:
matches: list[dict[str, Any]] = []
for row in rows:
score = _match_score(profile, row)
if score >= 3:
row_with_score = dict(row)
row_with_score["_match_score"] = score
matches.append(row_with_score)
matches.sort(key=lambda row: int(row.get("_match_score", 0)), reverse=True)
return matches
def _match_score(profile: dict[str, Any], row: dict[str, Any]) -> int:
score = 0
for field in COMPARATOR_FIELDS:
profile_value = profile.get(field)
row_value = row.get(field)
if profile_value is None or profile_value == "" or row_value == "":
continue
if field == "number_of_arms":
if _to_float(profile_value) == _to_float(row_value):
score += 1
continue
if str(profile_value).upper() == str(row_value).upper():
score += 1
return score
def _match_strategy(match_count: int) -> str:
if match_count >= 20:
return "matched_pubtime_prepared_fields"
return "domain_fallback_too_few_matches"
def _summarize_rows(rows: list[dict[str, Any]] | tuple[dict[str, Any], ...]) -> dict[str, Any]:
enrollments = [_to_float(row.get("enrollment")) for row in rows]
durations = [_to_float(row.get("actual_duration")) for row in rows]
facilities = [_to_float(row.get("number_of_facilities")) for row in rows]
arms = [_to_float(row.get("number_of_arms")) for row in rows]
primary_outcomes = [_to_float(row.get("number_of_primary_outcomes_to_measure")) for row in rows]
secondary_outcomes = [_to_float(row.get("number_of_secondary_outcomes_to_measure")) for row in rows]
time_to_pub = [_to_float(row.get("time_to_pub")) for row in rows]
return {
"median_enrollment": _median(enrollments),
"median_duration_months": _median(durations),
"median_facilities": _median(facilities),
"median_arms": _median(arms),
"median_primary_outcomes": _median(primary_outcomes),
"median_secondary_outcomes": _median(secondary_outcomes),
"median_time_to_publication_days": _median(time_to_pub),
"publication_rate": _rate(rows, _has_publication),
"results_reported_rate": _rate(rows, lambda row: _is_true(row.get("were_results_reported"))),
"dmc_rate": _rate(rows, lambda row: _is_true(row.get("has_dmc"))),
"us_facility_rate": _rate(rows, lambda row: _is_true(row.get("has_us_facility"))),
}
def _compare_profile(
profile: dict[str, Any], rows: list[dict[str, Any]] | tuple[dict[str, Any], ...]
) -> dict[str, Any]:
summary = _summarize_rows(rows)
flags: list[str] = []
enrollment = _to_float(profile.get("enrollment"))
median_enrollment = summary["median_enrollment"]
if enrollment and median_enrollment:
if enrollment >= median_enrollment * 1.5:
flags.append("Planned enrollment is substantially above the comparator median.")
elif enrollment <= median_enrollment * 0.5:
flags.append("Planned enrollment is substantially below the comparator median.")
facilities = _to_float(profile.get("number_of_facilities"))
median_facilities = summary["median_facilities"]
if facilities and median_facilities and facilities >= max(5, median_facilities * 2):
flags.append("Planned site count is high relative to comparators.")
arms = _to_float(profile.get("number_of_arms"))
median_arms = summary["median_arms"]
if arms and median_arms and arms >= max(4, median_arms * 2):
flags.append("Planned arm count is high relative to comparators.")
primary_outcomes = _to_float(profile.get("number_of_primary_outcomes"))
median_primary = summary["median_primary_outcomes"]
if primary_outcomes and median_primary and primary_outcomes >= max(3, median_primary * 2):
flags.append("Primary outcome count is high relative to comparators.")
secondary_outcomes = _to_float(profile.get("number_of_secondary_outcomes"))
median_secondary = summary["median_secondary_outcomes"]
if secondary_outcomes and median_secondary and secondary_outcomes >= max(6, median_secondary * 2):
flags.append("Secondary outcome count is high relative to comparators.")
dmc_rate = summary["dmc_rate"]
if profile.get("has_dmc") is False and dmc_rate is not None and dmc_rate >= 0.5:
flags.append("Most comparable trials have a DMC; this design does not.")
if profile.get("allocation") == "RANDOMIZED" and profile.get("masking") == "NONE":
flags.append("Randomized open-label design should be reviewed for bias and ascertainment risk.")
if profile.get("criteria") and len(str(profile["criteria"]).split()) < 20:
flags.append("Eligibility criteria are too short for strong complexity assessment.")
priority = "standard"
if len(flags) >= 3:
priority = "high"
elif flags:
priority = "focused"
return {
"review_priority": priority,
"flags": flags,
"note": "This is a PubTime-prepared historical comparator, not a validated prediction model.",
}
def _example_rows(
profile: dict[str, Any], rows: list[dict[str, Any]] | tuple[dict[str, Any], ...]
) -> list[dict[str, Any]]:
examples: list[dict[str, Any]] = []
ranked = sorted(rows, key=lambda row: _match_score(profile, row), reverse=True)
for row in ranked[:5]:
examples.append(
{
"nct_id": row.get("nct_id"),
"brief_title": row.get("brief_title"),
"phase": row.get("phase"),
"enrollment": _to_float(row.get("enrollment")),
"facilities": _to_float(row.get("number_of_facilities")),
"arms": _to_float(row.get("number_of_arms")),
"published": _has_publication(row),
}
)
return examples
def _median(values: list[float | None]) -> float | None:
cleaned = [value for value in values if value is not None]
if not cleaned:
return None
return round(float(pd.Series(cleaned).median()), 2)
def _rate(rows: list[dict[str, Any]] | tuple[dict[str, Any], ...], predicate) -> float | None:
if not rows:
return None
return round(sum(1 for row in rows if predicate(row)) / len(rows), 3)
def _to_float(value: Any) -> float | None:
if value is None or value == "":
return None
try:
return float(value)
except (TypeError, ValueError):
return None
def _is_true(value: Any) -> bool:
return str(value).strip().lower() == "true"
def _has_publication(row: dict[str, Any]) -> bool:
if row.get("event_pub") != "":
return _to_float(row.get("event_pub")) == 1
if row.get("pub_date"):
return True
result_count = _to_float(row.get("result_count"))
return bool(result_count and result_count > 0)
|