Datasets:
row_id
string | series_id
string | timepoint_h
int64 | host_model
string | drug
string | drug_conc_mg_L
float64 | expected_drug_conc_mg_L
float64 | conc_deviation
float64 | egfr_ml_min_1p73
float64 | expected_egfr_ml_min_1p73
float64 | egfr_drop_vs_expected
float64 | creatinine_mg_dL
float64 | expected_creatinine_mg_dL
float64 | creatinine_rise_vs_expected
float64 | renal_coherence_index
float64 | exposure_index
float64 | stress_index
float64 | later_severe_aki_flag
int64 | assay_method
string | source_type
string | renal_clearance_decoupling_signal
int64 | earliest_renal_clearance_decoupling
int64 | notes
string |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
ABXHT002-TR-0001
|
S1
| 0
|
human_sim
|
vancomycin
| 12
| 12
| 0
| 95
| 95
| 0
| 0.9
| 0.9
| 0
| 0.9
| 0.1
| 0.1
| 0
|
pk_creat_egfr_panel
|
simulated
| 0
| 0
|
baseline
|
ABXHT002-TR-0002
|
S1
| 12
|
human_sim
|
vancomycin
| 14
| 14
| 0
| 92
| 92
| 0
| 0.95
| 0.95
| 0
| 0.88
| 0.85
| 0.9
| 0
|
pk_creat_egfr_panel
|
simulated
| 0
| 0
|
tracks expected
|
ABXHT002-TR-0003
|
S1
| 24
|
human_sim
|
vancomycin
| 14.5
| 14
| 0.5
| 80
| 90
| 10
| 1.2
| 1
| 0.2
| 0.35
| 0.9
| 0.9
| 0
|
pk_creat_egfr_panel
|
simulated
| 0
| 0
|
first decoupling
|
ABXHT002-TR-0004
|
S1
| 36
|
human_sim
|
vancomycin
| 15
| 14
| 1
| 78
| 90
| 12
| 1.3
| 1
| 0.3
| 0.3
| 0.9
| 0.9
| 0
|
pk_creat_egfr_panel
|
simulated
| 1
| 1
|
confirmed onset
|
ABXHT002-TR-0005
|
S1
| 72
|
human_sim
|
vancomycin
| 20
| 15
| 5
| 55
| 88
| 33
| 2.1
| 1.05
| 1.05
| 0.2
| 0.9
| 0.9
| 1
|
pk_creat_egfr_panel
|
simulated
| 1
| 0
|
severe later
|
ABXHT002-TR-0006
|
S2
| 0
|
human_sim
|
gentamicin
| 4
| 4
| 0
| 100
| 100
| 0
| 0.8
| 0.8
| 0
| 0.9
| 0.1
| 0.1
| 0
|
pk_creat_egfr_panel
|
simulated
| 0
| 0
|
baseline
|
ABXHT002-TR-0007
|
S2
| 24
|
human_sim
|
gentamicin
| 4
| 4
| 0
| 98
| 98
| 0
| 0.82
| 0.82
| 0
| 0.86
| 0.9
| 0.9
| 0
|
pk_creat_egfr_panel
|
simulated
| 0
| 0
|
stable
|
ABXHT002-TR-0008
|
S2
| 48
|
human_sim
|
gentamicin
| 4
| 4
| 0
| 97
| 97
| 0
| 0.83
| 0.83
| 0
| 0.84
| 0.9
| 0.9
| 0
|
pk_creat_egfr_panel
|
simulated
| 0
| 0
|
stable
|
ABXHT002-TR-0009
|
S3
| 0
|
human_sim
|
vancomycin
| 14
| 14
| 0
| 35
| 35
| 0
| 2.2
| 2.2
| 0
| 0.9
| 0.9
| 0.9
| 1
|
pk_creat_egfr_panel
|
simulated
| 0
| 0
|
baseline renal poor
|
ABXHT002-TR-0010
|
S4
| 0
|
human_sim
|
vancomycin
| 15
| 14
| 1
| 78
| 90
| 12
| 1.3
| 1
| 0.3
| 0.3
| 0.9
| 0.3
| 1
|
pk_creat_egfr_panel
|
simulated
| 0
| 0
|
stress low
|
ABX-HT-002 Renal Clearance Decoupling
Purpose
Detect early kidney function marker drift that no longer tracks drug levels.
Core pattern
- stress_index high
- exposure_index high
- renal_coherence_index drops
- egfr_drop_vs_expected or creatinine_rise_vs_expected stays high
- later_severe_aki_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
- drug
- drug_conc_mg_L
- expected_drug_conc_mg_L
- conc_deviation
- egfr_ml_min_1p73
- expected_egfr_ml_min_1p73
- egfr_drop_vs_expected
- creatinine_mg_dL
- expected_creatinine_mg_dL
- creatinine_rise_vs_expected
- renal_coherence_index
- exposure_index
- stress_index
- later_severe_aki_flag
- assay_method
- source_type
- renal_clearance_decoupling_signal
- earliest_renal_clearance_decoupling
Labels
renal_clearance_decoupling_signal
- 1 for rows at or after first confirmed decoupling onset
earliest_renal_clearance_decoupling
- 1 only for the first onset row in that series
Scorer logic in v1
- exclude series with low baseline eGFR
- candidate onset point
- stress_index at least 0.80
- exposure_index at least 0.80
- renal_coherence_index at most 0.40
- egfr_drop_vs_expected at least 10 or creatinine_rise_vs_expected at least 0.20
- for two consecutive points
- exclude one point extreme drop then recovery artifacts
- confirmation
- later_severe_aki_flag equals 1 later in series
Evaluation
Run
- python scorer.py --path data/test.csv
- Downloads last month
- 17