CoolFace
Datasetpublic

ClarusC64/ABX-CT-008_toxicity_additivity_warning-v0.1

ABX-CT-008 Toxicity-Additivity Warning Purpose Detect when combined toxicity exceeds simple additivity from each drug alone. Core pattern stress_index high exposure indices high tox_excess rises and stays high later severe_toxicity_flag appears Files data/train.csv data/test.csv scorer.py Schema Each row is one timepoint in a within series time course. Required columns row_id series_id timepoint_h host_model tox_biomarker_name tox_biomarker_units stress_index exposure_a_index… See the full description on the dataset page: https://huggingface.co/datasets/ClarusC64/ABX-CT-008_toxicity_additivity_warning-v0.1.

sourceHugging Facemitupdated 8mo agoView on Hugging Face
0likes22downloads
Dataset Card

ABX-CT-008 Toxicity-Additivity Warning

Purpose

Detect when combined toxicity exceeds simple additivity from each drug alone.

Core pattern

  • —stress_index high
  • —exposure indices high
  • —tox_excess rises and stays high
  • —later severetoxicityflag appears

Files

  • —data/train.csv
  • —data/test.csv
  • —scorer.py

Schema

Each row is one timepoint in a within series time course.

Required columns

  • —row_id
  • —series_id
  • —timepoint_h
  • —host_model
  • —toxbiomarkername
  • —toxbiomarkerunits
  • —stress_index
  • —exposureaindex
  • —exposurebindex
  • —baselinetoxbiomarker
  • —monoAtoxdelta
  • —monoBtoxdelta
  • —expectedadditivetox
  • —observedcombotox
  • —tox_excess
  • —severetoxicityflag
  • —drug_a
  • —drug_b
  • —source_type
  • —toxicityadditivitywarning_signal
  • —earliesttoxicityadditivity_warning

Labels

  • —toxicityadditivitywarning_signal
  • —1 for rows at or after first confirmed warning point
  • —earliesttoxicityadditivity_warning
  • —1 only for the first warning row in that series

Scorer logic in v1

  • —baselinetoxbiomarker must be below a simple ceiling
  • —candidate warning point
  • —stress_index at least 0.80
  • —exposureaindex and exposurebindex at least 0.80
  • —tox_excess at least 0.40 at two consecutive timepoints
  • —exclude tox_excess spike then snapback artifacts
  • —confirmation
  • —severetoxicityflag equals 1 later in series

Evaluation

Run

  • —python scorer.py --path data/test.csv