Dataset Viewer
Auto-converted to Parquet Duplicate
row_id
string
series_id
string
timepoint_h
int64
host_model
string
drug
string
drug_conc_ng_mL
int64
expected_drug_conc_ng_mL
int64
conc_deviation
int64
qtc_ms
int64
expected_qtc_ms
int64
qtc_rise_vs_expected_ms
int64
cardiac_coherence_index
float64
exposure_index
float64
stress_index
float64
later_serious_arrhythmia_flag
int64
assay_method
string
source_type
string
qt_risk_signal
int64
earliest_qt_risk
int64
notes
string
ABXHT004-TR-0001
S1
0
human_sim
azithromycin
80
80
0
410
410
0
0.9
0.1
0.1
0
ecg_pk_panel
simulated
0
0
baseline
ABXHT004-TR-0002
S1
6
human_sim
azithromycin
120
120
0
415
415
0
0.88
0.85
0.9
0
ecg_pk_panel
simulated
0
0
tracks expected
ABXHT004-TR-0003
S1
12
human_sim
azithromycin
125
120
5
450
420
30
0.35
0.9
0.9
0
ecg_pk_panel
simulated
0
0
first decoupling
ABXHT004-TR-0004
S1
18
human_sim
azithromycin
130
120
10
460
420
40
0.3
0.9
0.9
0
ecg_pk_panel
simulated
1
1
confirmed onset
ABXHT004-TR-0005
S1
48
human_sim
azithromycin
140
125
15
520
425
95
0.2
0.9
0.9
1
ecg_pk_panel
simulated
1
0
arrhythmia later
ABXHT004-TR-0006
S2
0
human_sim
levofloxacin
90
90
0
405
405
0
0.9
0.1
0.1
0
ecg_pk_panel
simulated
0
0
baseline
ABXHT004-TR-0007
S2
12
human_sim
levofloxacin
110
110
0
410
410
0
0.86
0.9
0.9
0
ecg_pk_panel
simulated
0
0
stable
ABXHT004-TR-0008
S2
24
human_sim
levofloxacin
110
110
0
412
412
0
0.84
0.9
0.9
0
ecg_pk_panel
simulated
0
0
stable
ABXHT004-TR-0009
S3
0
human_sim
azithromycin
120
120
0
490
490
0
0.9
0.9
0.9
1
ecg_pk_panel
simulated
0
0
baseline QTc high
ABXHT004-TR-0010
S4
0
human_sim
azithromycin
130
120
10
460
420
40
0.3
0.9
0.3
1
ecg_pk_panel
simulated
0
0
stress low

ABX-HT-004 QT Prolongation Prediction

Purpose

Detect early QT risk when cardiac repolarization markers stop tracking drug levels before serious arrhythmia events.

Core pattern

  • stress_index high
  • exposure_index high
  • cardiac_coherence_index drops
  • qtc_rise_vs_expected_ms stays high
  • later_serious_arrhythmia_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_ng_mL
  • expected_drug_conc_ng_mL
  • conc_deviation
  • qtc_ms
  • expected_qtc_ms
  • qtc_rise_vs_expected_ms
  • cardiac_coherence_index
  • exposure_index
  • stress_index
  • later_serious_arrhythmia_flag
  • assay_method
  • source_type
  • qt_risk_signal
  • earliest_qt_risk

Labels

  • qt_risk_signal

    • 1 for rows at or after first confirmed QT risk onset
  • earliest_qt_risk

    • 1 only for the first onset row in that series

Scorer logic in v1

  • exclude series with high baseline QTc
  • candidate onset point
    • stress_index at least 0.80
    • exposure_index at least 0.80
    • cardiac_coherence_index at most 0.40
    • qtc_rise_vs_expected_ms at least 25 for two consecutive points
    • ignore one point QTc spike then recovery artifacts
  • confirmation
    • later_serious_arrhythmia_flag equals 1 later in series

Evaluation

Run

  • python scorer.py --path data/test.csv
Downloads last month
12