Datasets:
row_id string | series_id string | treatment_day int64 | host_model string | drug string | drug_exposure_index float64 | uv_index int64 | uv_exposure_hours float64 | uv_dose_index float64 | expected_uv_dose_index float64 | uv_dose_deviation float64 | skin_reaction_score float64 | expected_skin_reaction_score float64 | reaction_rise_vs_expected float64 | phototox_coherence_index float64 | host_stress_index float64 | later_severe_phototox_flag int64 | assay_method string | source_type string | phototoxicity_signal int64 | earliest_phototoxicity int64 | notes string |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
ABXHT008-TR-0001 | S1 | 0 | human_sim | doxycycline | 0.1 | 2 | 0.5 | 0.05 | 0.05 | 0 | 0 | 0 | 0 | 0.9 | 0.1 | 0 | uv_skin_panel | simulated | 0 | 0 | baseline |
ABXHT008-TR-0002 | S1 | 3 | human_sim | doxycycline | 0.9 | 6 | 2 | 0.55 | 0.55 | 0 | 1 | 1 | 0 | 0.88 | 0.9 | 0 | uv_skin_panel | simulated | 0 | 0 | tracks expected |
ABXHT008-TR-0003 | S1 | 4 | human_sim | doxycycline | 0.9 | 6 | 2 | 0.55 | 0.55 | 0 | 4.5 | 1.2 | 3.3 | 0.35 | 0.9 | 0 | uv_skin_panel | simulated | 0 | 0 | first divergence |
ABXHT008-TR-0004 | S1 | 5 | human_sim | doxycycline | 0.9 | 6 | 2 | 0.55 | 0.55 | 0 | 5 | 1.2 | 3.8 | 0.3 | 0.9 | 0 | uv_skin_panel | simulated | 1 | 1 | onset |
ABXHT008-TR-0005 | S1 | 7 | human_sim | doxycycline | 0.9 | 7 | 3 | 0.8 | 0.75 | 0.05 | 8.5 | 2 | 6.5 | 0.2 | 0.9 | 1 | uv_skin_panel | simulated | 1 | 0 | severe later |
ABXHT008-TR-0006 | S2 | 0 | human_sim | levofloxacin | 0.1 | 2 | 0.5 | 0.05 | 0.05 | 0 | 0 | 0 | 0 | 0.9 | 0.1 | 0 | uv_skin_panel | simulated | 0 | 0 | baseline |
ABXHT008-TR-0007 | S2 | 3 | human_sim | levofloxacin | 0.9 | 5 | 1.5 | 0.4 | 0.4 | 0 | 1.2 | 1.1 | 0.1 | 0.86 | 0.9 | 0 | uv_skin_panel | simulated | 0 | 0 | stable |
ABXHT008-TR-0008 | S2 | 6 | human_sim | levofloxacin | 0.9 | 5 | 1.5 | 0.4 | 0.4 | 0 | 1 | 1.1 | -0.1 | 0.84 | 0.9 | 0 | uv_skin_panel | simulated | 0 | 0 | stable |
ABXHT008-TR-0009 | S3 | 0 | human_sim | doxycycline | 0.9 | 6 | 2 | 0.55 | 0.55 | 0 | 5 | 1.2 | 3.8 | 0.3 | 0.9 | 1 | uv_skin_panel | simulated | 0 | 0 | baseline reaction high |
ABXHT008-TR-0010 | S4 | 5 | human_sim | doxycycline | 0.9 | 6 | 2 | 0.55 | 0.55 | 0 | 5 | 1.2 | 3.8 | 0.3 | 0.3 | 1 | uv_skin_panel | simulated | 0 | 0 | stress low |
ABX-HT-008 Phototoxicity Activation
Purpose
Detect early phototoxicity activation when sunlight interaction stops behaving predictably.
Core pattern
- host_stress_index high
- drug_exposure_index high
- uv_dose_index high
- phototox_coherence_index drops
- reaction_rise_vs_expected stays high
- later_severe_phototox_flag appears
Files
- data/train.csv
- data/test.csv
- scorer.py
Schema
Each row is one timepoint in a within series treatment course.
Required columns
- row_id
- series_id
- treatment_day
- host_model
- drug
- drug_exposure_index
- uv_index
- uv_exposure_hours
- uv_dose_index
- expected_uv_dose_index
- uv_dose_deviation
- skin_reaction_score
- expected_skin_reaction_score
- reaction_rise_vs_expected
- phototox_coherence_index
- host_stress_index
- later_severe_phototox_flag
- assay_method
- source_type
- phototoxicity_signal
- earliest_phototoxicity
Labels
phototoxicity_signal
- 1 for rows at or after first confirmed activation onset
earliest_phototoxicity
- 1 only for the first onset row in that series
Scorer logic in v1
- exclude series with high baseline skin reaction
- candidate onset point
- host_stress_index at least 0.80
- drug_exposure_index at least 0.80
- uv_dose_index at least 0.35
- phototox_coherence_index at most 0.40
- reaction_rise_vs_expected at least 2.5 for two consecutive points
- ignore one point reaction spike then recovery artifacts
- confirmation
- later_severe_phototox_flag equals 1 later in series
Evaluation
Run
- python scorer.py --path data/test.csv
- Downloads last month
- 11