Datasets:
metadata
license: mit
language:
- en
pretty_name: ABX-HT-010 Drug-Drug Interaction Emergence
tags:
- antibiotics
- toxicity
- ddi
- drug-interaction
- pharmacokinetics
- tabular
task_categories:
- tabular-classification
size_categories:
- n<1k
ABX-HT-010 Drug-Drug Interaction Emergence
Purpose
Detect early interaction emergence when antibiotic exposure stops tracking expected interaction shift with concomitant medications.
Core pattern
- host_stress_index high
- concomitant_exposure_index high
- ddi_coherence_index drops
- interaction_shift_deviation stays high
- later_adverse_ddi_flag appears
Files
- data/train.csv
- data/test.csv
- scorer.py
Schema
Each row is one timepoint in a within series PK and medication exposure course.
Required columns
- row_id
- series_id
- timepoint_h
- host_model
- abx_drug
- abx_dose_mg
- abx_conc_ng_mL
- expected_abx_conc_ng_mL
- abx_conc_deviation
- concomitant_drug
- concomitant_exposure_index
- expected_interaction_shift
- interaction_shift_deviation
- ddi_coherence_index
- host_stress_index
- later_adverse_ddi_flag
- assay_method
- source_type
- ddi_emergence_signal
- earliest_ddi_emergence
Labels
ddi_emergence_signal
- 1 for rows at or after first confirmed emergence onset
earliest_ddi_emergence
- 1 only for the first onset row in that series
Scorer logic in v1
- require a concomitant medication signal in the series
- candidate onset point
- host_stress_index at least 0.80
- concomitant_exposure_index at least 0.80
- ddi_coherence_index at most 0.40
- absolute interaction_shift_deviation at least 40 for two consecutive points
- ignore one point deviation spike then recovery artifacts
- confirmation
- later_adverse_ddi_flag equals 1 later in series
Evaluation
Run
- python scorer.py --path data/test.csv