Datasets:

ClarusC64
/

ABX-HT-006_microbiome_diversity_collapse-v0.1

row_id string	series_id string	timepoint_d int64	host_model string	drug string	abx_exposure_index float64	host_stress_index float64	shannon_diversity float64	expected_shannon_diversity float64	shannon_drop_vs_expected float64	evenness_index float64	expected_evenness_index float64	evenness_drop_vs_expected float64	domination_index float64	expected_domination_index float64	domination_rise_vs_expected float64	microbiome_coherence_index float64	later_cdiff_flag int64	assay_method string	source_type string	microbiome_collapse_signal int64	earliest_microbiome_collapse int64	notes string
ABXHT006-TR-0001	S1	0	human_sim	clindamycin	0.1	0.1	3.6	3.6	0	0.78	0.78	0	0.2	0.2	0	0.9	0	16s_panel	simulated	0	0	baseline
ABXHT006-TR-0002	S1	3	human_sim	clindamycin	0.9	0.9	3.4	3.45	0.05	0.75	0.76	0.01	0.22	0.22	0	0.86	0	16s_panel	simulated	0	0	expected response
ABXHT006-TR-0003	S1	7	human_sim	clindamycin	0.9	0.9	2.4	3.3	0.9	0.52	0.74	0.22	0.55	0.25	0.3	0.32	0	16s_panel	simulated	0	0	first collapse drift
ABXHT006-TR-0004	S1	10	human_sim	clindamycin	0.9	0.9	2.1	3.2	1.1	0.48	0.73	0.25	0.62	0.26	0.36	0.28	0	16s_panel	simulated	1	1	confirmed onset
ABXHT006-TR-0005	S1	21	human_sim	clindamycin	0.9	0.9	1.5	3	1.5	0.4	0.72	0.32	0.78	0.28	0.5	0.2	1	16s_panel	simulated	1	0	cdiff later
ABXHT006-TR-0006	S2	0	human_sim	amoxicillin	0.1	0.1	3.5	3.5	0	0.77	0.77	0	0.2	0.2	0	0.9	0	16s_panel	simulated	0	0	baseline
ABXHT006-TR-0007	S2	7	human_sim	amoxicillin	0.9	0.9	3.1	3.2	0.1	0.72	0.74	0.02	0.25	0.23	0.02	0.84	0	16s_panel	simulated	0	0	mild shift
ABXHT006-TR-0008	S2	14	human_sim	amoxicillin	0.9	0.9	3.2	3.1	-0.1	0.74	0.72	-0.02	0.22	0.24	-0.02	0.83	0	16s_panel	simulated	0	0	recovery
ABXHT006-TR-0009	S3	0	human_sim	clindamycin	0.9	0.9	1.8	1.8	0	0.5	0.5	0	0.6	0.6	0	0.9	1	16s_panel	simulated	0	0	baseline diversity low
ABXHT006-TR-0010	S4	0	human_sim	clindamycin	0.9	0.3	2.1	3.2	1.1	0.48	0.73	0.25	0.62	0.26	0.36	0.28	1	16s_panel	simulated	0	0	stress low

ABX-HT-006 Microbiome Diversity Collapse

Purpose

Detect early microbiome collapse that predicts C. difficile risk.

Core pattern

host_stress_index high
abx_exposure_index high
microbiome_coherence_index drops
shannon_drop_vs_expected stays high
domination_rise_vs_expected stays high
later_cdiff_flag appears

Files

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

Schema

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

Required columns

row_id
series_id
timepoint_d
host_model
drug
abx_exposure_index
host_stress_index
shannon_diversity
expected_shannon_diversity
shannon_drop_vs_expected
evenness_index
expected_evenness_index
evenness_drop_vs_expected
domination_index
expected_domination_index
domination_rise_vs_expected
microbiome_coherence_index
later_cdiff_flag
assay_method
source_type
microbiome_collapse_signal
earliest_microbiome_collapse

Labels

microbiome_collapse_signal
- 1 for rows at or after first confirmed collapse onset
earliest_microbiome_collapse
- 1 only for the first onset row in that series

Scorer logic in v1

exclude series with low baseline Shannon diversity
candidate onset point
- host_stress_index at least 0.80
- abx_exposure_index at least 0.80
- microbiome_coherence_index at most 0.40
- shannon_drop_vs_expected at least 0.80
- domination_rise_vs_expected at least 0.25
- for two consecutive points
- ignore one point diversity drop then recovery artifacts
confirmation
- later_cdiff_flag equals 1 later in series

Evaluation

Run

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

Downloads last month: 15

Size of downloaded dataset files:

3.72 kB

Size of the auto-converted Parquet files:

24.3 kB

Number of rows: