Datasets:

electricsheepafrica
/

ssa-breast-integrated-ml

sample_id stringlengths 23 23	population stringclasses 7 values	region stringclasses 5 values	is_SSA bool 2 classes	is_reference_panel bool 2 classes	sex stringclasses 2 values	age int64 18 90 ⌀	bmi float64 14 47.1 ⌀	ses stringclasses 3 values	prs_category stringclasses 3 values	prs_standardized float64 -2.99 2.55 ⌀	high_risk_germline bool 2 classes	stage_at_diagnosis stringclasses 4 values	grade int64 1 3 ⌀	tumour_size_cm float64 0.2 13.8 ⌀	treatment_regimen stringclasses 5 values	tumor_subtype stringclasses 5 values	er_status stringclasses 2 values	pr_status stringclasses 2 values	her2_status stringclasses 2 values	nlr_category stringclasses 3 values	triple_negative bool 2 classes	os_time_years float64 0 9.99 ⌀	os_event bool 2 classes
INTEGRATED_SAMPLE_00001	SSA_West	West	true	false	Female	53	25.2	High	Low	-0.192	false	III	3	7.33	Endocrine_only	Luminal_A	Positive	Positive	Negative	Low	false	6.208	false
INTEGRATED_SAMPLE_00002	SSA_West	West	true	false	Female	39	24.2	Low	Average	-0.403	false	IV	3	7.46	Palliative_only	Basal_like	Negative	Negative	Negative	Intermediate	true	2.356	false
INTEGRATED_SAMPLE_00003	SSA_West	West	true	false	Female	58	29.7	Low	High	-0.283	false	III	3	4.68	Chemo_plus_targeted	Basal_like	Negative	Negative	Negative	Intermediate	true	0.056	true
INTEGRATED_SAMPLE_00004	SSA_West	West	true	false	Female	60	28.8	Middle	Average	0.36	false	II	2	1.43	Chemo_only	Luminal_A	Positive	Positive	Positive	Intermediate	false	5.918	true
INTEGRATED_SAMPLE_00005	SSA_West	West	true	false	Female	29	20.5	Middle	Average	-0.925	false	II	2	2.32	Chemo_only	HER2_enriched	Positive	Positive	Positive	Intermediate	false	2.987	true
INTEGRATED_SAMPLE_00006	SSA_West	West	true	false	Female	36	33.1	Low	Average	0.458	false	I	3	2.01	Endocrine_only	HER2_enriched	Positive	Positive	Positive	Intermediate	false	4.679	true
INTEGRATED_SAMPLE_00007	SSA_West	West	true	false	Female	51	27	Middle	Average	0.19	false	IV	3	11.17	Endocrine_only	Basal_like	Negative	Negative	Negative	Low	true	0.09	true
INTEGRATED_SAMPLE_00008	SSA_West	West	true	false	Female	47	21.9	Low	High	0.759	false	II	2	4.73	Endocrine_only	Basal_like	Negative	Negative	Negative	Intermediate	true	0.268	true
INTEGRATED_SAMPLE_00009	SSA_West	West	true	false	Female	50	33.4	Middle	High	-0.179	false	IV	3	8.61	Chemo_only	Basal_like	Negative	Negative	Negative	Low	true	0.943	true
INTEGRATED_SAMPLE_00010	SSA_West	West	true	false	Female	41	30.6	Middle	Low	-0.779	false	IV	2	6.65	Chemo_plus_targeted	Basal_like	Negative	Negative	Negative	High	true	1.805	true
INTEGRATED_SAMPLE_00011	SSA_West	West	true	false	Female	60	21.2	Low	Low	-0.171	false	I	2	2.21	Chemo_only	Basal_like	Negative	Negative	Negative	Low	true	3.984	false
INTEGRATED_SAMPLE_00012	SSA_West	West	true	false	Female	59	36.4	High	Low	-0.503	false	III	3	7.14	Endocrine_only	Basal_like	Negative	Negative	Negative	High	true	0.343	true
INTEGRATED_SAMPLE_00013	SSA_West	West	true	false	Female	51	30.8	High	Average	1.488	false	II	2	1.85	Chemo_only	HER2_enriched	Positive	Positive	Positive	High	false	0.934	true
INTEGRATED_SAMPLE_00014	SSA_West	West	true	false	Female	62	27.4	Low	High	1.115	true	III	3	5.62	Chemo_only	Luminal_A	Positive	Positive	Negative	Low	false	0.74	true
INTEGRATED_SAMPLE_00015	SSA_West	West	true	false	Female	55	35.5	Low	Average	0.615	false	IV	3	8.82	Surgery_only	Luminal_A	Positive	Positive	Negative	High	false	0.847	false
INTEGRATED_SAMPLE_00016	SSA_West	West	true	false	Female	41	28.3	Middle	Average	-0.32	false	IV	2	9.75	Chemo_plus_targeted	Basal_like	Negative	Negative	Negative	Low	true	0.472	true
INTEGRATED_SAMPLE_00017	SSA_West	West	true	false	Female	54	27.4	Low	Average	0.406	false	III	2	7.91	Chemo_plus_targeted	Luminal_B	Positive	Positive	Negative	Low	false	0.228	true
INTEGRATED_SAMPLE_00018	SSA_West	West	true	false	Female	39	24.2	Low	High	0.259	false	III	3	5.65	Endocrine_only	Basal_like	Negative	Negative	Negative	Intermediate	true	3.832	true
INTEGRATED_SAMPLE_00019	SSA_West	West	true	false	Female	60	30.4	Low	Low	-0.719	false	II	2	2.76	Endocrine_only	Luminal_B	Positive	Positive	Negative	High	false	5.219	false
INTEGRATED_SAMPLE_00020	SSA_West	West	true	false	Female	49	29.9	High	Average	-0.968	false	II	2	1.82	Chemo_only	Luminal_B	Positive	Positive	Negative	Intermediate	false	0.589	true
INTEGRATED_SAMPLE_00021	SSA_West	West	true	false	Female	48	27.6	Low	Average	0.098	false	III	3	6.66	Chemo_plus_targeted	Basal_like	Negative	Negative	Negative	Intermediate	true	6.83	true
INTEGRATED_SAMPLE_00022	SSA_West	West	true	false	Female	43	23.6	Middle	Average	-0.091	false	I	2	2.74	Chemo_plus_targeted	Luminal_B	Positive	Positive	Negative	Intermediate	false	5.739	false
INTEGRATED_SAMPLE_00023	SSA_West	West	true	false	Female	63	28.1	Low	Average	0.346	false	III	1	6.81	Chemo_plus_targeted	Basal_like	Negative	Negative	Negative	High	true	0.502	true
INTEGRATED_SAMPLE_00024	SSA_West	West	true	false	Female	48	27.2	Low	Average	-0.599	false	IV	1	6.76	Palliative_only	Luminal_A	Positive	Positive	Positive	Low	false	0.207	true
INTEGRATED_SAMPLE_00025	SSA_West	West	true	false	Female	45	20.8	High	High	0.579	false	III	3	6.35	Chemo_plus_targeted	Basal_like	Negative	Negative	Negative	Intermediate	true	0.305	true
INTEGRATED_SAMPLE_00026	SSA_West	West	true	false	Female	46	20.1	Low	Average	0.4	true	II	2	5.11	Chemo_only	Luminal_A	Positive	Positive	Negative	High	false	8.463	false
INTEGRATED_SAMPLE_00027	SSA_West	West	true	false	Female	56	30.3	Low	High	1.135	false	III	2	4.83	Chemo_only	Basal_like	Negative	Negative	Negative	High	true	0.034	true
INTEGRATED_SAMPLE_00028	SSA_West	West	true	false	Female	54	14.5	Low	Average	-0.264	false	III	2	6.57	Chemo_only	Basal_like	Negative	Negative	Negative	Intermediate	true	0.207	true
INTEGRATED_SAMPLE_00029	SSA_West	West	true	false	Female	55	28.5	Middle	High	0.336	false	II	2	7.63	Palliative_only	Luminal_B	Positive	Positive	Negative	Intermediate	false	0.861	false
INTEGRATED_SAMPLE_00030	SSA_West	West	true	false	Female	55	25.1	Low	Low	-0.181	false	II	1	6.94	Surgery_only	Luminal_A	Positive	Positive	Negative	High	false	1.596	true
INTEGRATED_SAMPLE_00031	SSA_West	West	true	false	Male	74	34	Low	Average	-1.093	false	III	3	4.48	Chemo_plus_targeted	Basal_like	Negative	Negative	Negative	Low	true	2.965	true
INTEGRATED_SAMPLE_00032	SSA_West	West	true	false	Female	46	25	Low	Low	-0.644	false	IV	3	6.57	Palliative_only	Basal_like	Negative	Negative	Negative	High	true	0.008	true
INTEGRATED_SAMPLE_00033	SSA_West	West	true	false	Female	44	30.8	Low	Average	-0.101	false	II	2	4.05	Chemo_plus_targeted	Luminal_B	Positive	Positive	Negative	Intermediate	false	0.016	true
INTEGRATED_SAMPLE_00034	SSA_West	West	true	false	Female	41	25.5	Low	Low	-0.916	false	III	3	6.87	Chemo_only	Luminal_A	Positive	Negative	Negative	Intermediate	false	1.474	true
INTEGRATED_SAMPLE_00035	SSA_West	West	true	false	Female	57	28.1	Middle	Low	-0.592	false	III	2	7.82	Chemo_plus_targeted	Basal_like	Negative	Negative	Negative	High	true	0.435	true
INTEGRATED_SAMPLE_00036	SSA_West	West	true	false	Female	62	31.2	Low	Average	0.254	false	IV	3	9.65	Chemo_plus_targeted	HER2_enriched	Positive	Negative	Positive	Intermediate	false	0.611	true
INTEGRATED_SAMPLE_00037	SSA_West	West	true	false	Female	49	25.2	Low	Low	-0.514	false	I	1	3.76	Surgery_only	Basal_like	Negative	Negative	Negative	Intermediate	true	1.637	false
INTEGRATED_SAMPLE_00038	SSA_West	West	true	false	Female	41	32.5	Middle	Average	1.206	false	III	3	6.91	Chemo_only	Luminal_A	Negative	Positive	Positive	Intermediate	false	4.252	true
INTEGRATED_SAMPLE_00039	SSA_West	West	true	false	Female	41	22.5	Middle	High	1.129	false	IV	2	7.53	Palliative_only	Basal_like	Negative	Negative	Negative	Low	true	1.513	true
INTEGRATED_SAMPLE_00040	SSA_West	West	true	false	Female	57	29	Low	Low	-0.682	false	II	1	6.05	Chemo_only	Luminal_B	Positive	Positive	Negative	Intermediate	false	2.07	false
INTEGRATED_SAMPLE_00041	SSA_West	West	true	false	Female	58	32.1	Low	Low	-0.948	false	II	2	6.68	Chemo_only	Luminal_A	Positive	Positive	Negative	Intermediate	false	7.299	false
INTEGRATED_SAMPLE_00042	SSA_West	West	true	false	Female	56	30.8	Low	Low	-0.891	false	I	2	0.89	Chemo_only	HER2_enriched	Negative	Negative	Positive	High	false	1.696	false
INTEGRATED_SAMPLE_00043	SSA_West	West	true	false	Female	43	37.1	High	Average	0.931	false	II	2	3.65	Endocrine_only	HER2_enriched	Positive	Negative	Positive	Low	false	2.014	true
INTEGRATED_SAMPLE_00044	SSA_West	West	true	false	Female	53	23.7	Low	Average	0.182	false	IV	3	9.21	Palliative_only	Luminal_B	Positive	Positive	Positive	Low	false	0.017	true
INTEGRATED_SAMPLE_00045	SSA_West	West	true	false	Female	51	20.6	Low	Low	0.115	false	II	3	3.77	Chemo_only	Basal_like	Negative	Negative	Positive	Low	false	7.648	false
INTEGRATED_SAMPLE_00046	SSA_West	West	true	false	Female	52	22.5	Low	Average	-0.43	false	III	3	4.18	Chemo_plus_targeted	Luminal_A	Positive	Positive	Negative	Low	false	3.62	false
INTEGRATED_SAMPLE_00047	SSA_West	West	true	false	Female	60	30.8	Low	Average	-0.068	false	II	3	5.27	Chemo_only	Luminal_B	Positive	Positive	Negative	Low	false	1.325	true
INTEGRATED_SAMPLE_00048	SSA_West	West	true	false	Female	52	30	Low	Average	-0.171	true	III	3	5.89	Chemo_plus_targeted	Basal_like	Negative	Negative	Negative	Intermediate	true	1.912	true
INTEGRATED_SAMPLE_00049	SSA_West	West	true	false	Female	57	28	Low	High	0.757	false	III	3	6.1	Chemo_plus_targeted	Basal_like	Negative	Negative	Negative	Low	true	0.411	true
INTEGRATED_SAMPLE_00050	SSA_West	West	true	false	Female	51	21.9	Middle	Average	0.84	false	III	2	7.89	Endocrine_only	Luminal_A	Positive	Positive	Negative	Intermediate	false	3.258	true
INTEGRATED_SAMPLE_00051	SSA_West	West	true	false	Female	53	31.3	Middle	High	1	false	II	2	3.18	Surgery_only	Basal_like	Negative	Negative	Negative	Low	true	6.162	false
INTEGRATED_SAMPLE_00052	SSA_West	West	true	false	Female	57	21.4	Middle	Low	-0.289	false	III	3	7.08	Chemo_only	Basal_like	Positive	Negative	Positive	Intermediate	false	0.513	true
INTEGRATED_SAMPLE_00053	SSA_West	West	true	false	Female	34	33.3	Middle	Average	0.905	false	III	1	0.87	Surgery_only	Basal_like	Negative	Positive	Negative	Intermediate	false	1.805	true
INTEGRATED_SAMPLE_00054	SSA_West	West	true	false	Female	46	41.8	Low	Low	-1.037	false	II	2	3.46	Chemo_plus_targeted	Basal_like	Negative	Negative	Negative	Intermediate	true	2.729	false
INTEGRATED_SAMPLE_00055	SSA_West	West	true	false	Female	45	29.3	High	Average	0.006	false	III	3	5.61	Chemo_only	Luminal_A	Positive	Positive	Negative	High	false	1.576	true
INTEGRATED_SAMPLE_00056	SSA_West	West	true	false	Female	43	30.4	Middle	Average	0.852	false	II	2	8.12	Endocrine_only	Luminal_A	Positive	Positive	Negative	Low	false	5.841	true
INTEGRATED_SAMPLE_00057	SSA_West	West	true	false	Female	47	25	Middle	Low	-0.127	false	III	3	4.44	Surgery_only	Basal_like	Negative	Negative	Negative	Intermediate	true	5.01	false
INTEGRATED_SAMPLE_00058	SSA_West	West	true	false	Female	66	18.4	Middle	Average	0.845	false	I	2	0.2	Surgery_only	Basal_like	Negative	Negative	Negative	Intermediate	true	5.576	false
INTEGRATED_SAMPLE_00059	SSA_West	West	true	false	Female	40	24.5	Low	High	0.727	false	I	1	2.49	Surgery_only	Basal_like	Negative	Negative	Negative	Intermediate	true	6.17	false
INTEGRATED_SAMPLE_00060	SSA_West	West	true	false	Female	61	22	Low	Average	0.371	false	I	2	1.88	Surgery_only	Luminal_A	Positive	Positive	Negative	High	false	6.602	false
INTEGRATED_SAMPLE_00061	SSA_West	West	true	false	Female	31	20.3	Low	High	1.612	false	I	1	4.49	Surgery_only	HER2_enriched	Negative	Positive	Positive	Intermediate	false	9.755	false
INTEGRATED_SAMPLE_00062	SSA_West	West	true	false	Female	46	36.1	Low	Average	-0.45	false	III	3	5.06	Palliative_only	Basal_like	Negative	Negative	Negative	High	true	0.019	true
INTEGRATED_SAMPLE_00063	SSA_West	West	true	false	Female	52	28.9	Middle	Average	0.252	false	IV	3	5.46	Palliative_only	Basal_like	Negative	Positive	Negative	Low	false	1.002	true
INTEGRATED_SAMPLE_00064	SSA_West	West	true	false	Female	56	29.4	Low	Low	-0.62	false	IV	3	3.92	Chemo_plus_targeted	Basal_like	Negative	Negative	Negative	Intermediate	true	0.262	true
INTEGRATED_SAMPLE_00065	SSA_West	West	true	false	Female	58	26	Low	High	0.978	false	III	2	3.45	Chemo_only	Basal_like	Negative	Negative	Negative	Intermediate	true	0.918	true
INTEGRATED_SAMPLE_00066	SSA_West	West	true	false	Female	59	26.6	High	Low	0.138	false	I	1	4.6	Surgery_only	Luminal_A	Positive	Positive	Positive	Low	false	4.533	false
INTEGRATED_SAMPLE_00067	SSA_West	West	true	false	Female	46	32.3	Middle	Average	-0.043	false	II	2	7.23	Chemo_only	HER2_enriched	Negative	Positive	Positive	Intermediate	false	6.101	true
INTEGRATED_SAMPLE_00068	SSA_West	West	true	false	Female	45	27.3	Middle	Low	-0.166	false	III	3	9.8	Chemo_only	Luminal_B	Positive	Positive	Negative	Intermediate	false	6.287	false
INTEGRATED_SAMPLE_00069	SSA_West	West	true	false	Female	59	30.2	Low	High	-0.228	false	IV	3	7.21	Chemo_plus_targeted	Luminal_A	Positive	Positive	Negative	Intermediate	false	0.094	true
INTEGRATED_SAMPLE_00070	SSA_West	West	true	false	Female	48	26.5	Middle	Average	-0.001	false	II	2	5.6	Endocrine_only	Basal_like	Positive	Negative	Negative	Intermediate	false	0.614	true
INTEGRATED_SAMPLE_00071	SSA_West	West	true	false	Female	36	25.8	High	Average	-0.043	false	III	3	9.15	Endocrine_only	Luminal_A	Positive	Positive	Negative	Intermediate	false	0.984	true
INTEGRATED_SAMPLE_00072	SSA_West	West	true	false	Female	38	20.5	Middle	Average	-1.24	false	II	3	3.63	Surgery_only	Basal_like	Negative	Negative	Negative	High	true	1.448	false
INTEGRATED_SAMPLE_00073	SSA_West	West	true	false	Female	40	17.6	Middle	Average	-0.798	false	IV	3	10.46	Palliative_only	Basal_like	Negative	Negative	Negative	Low	true	0.106	true
INTEGRATED_SAMPLE_00074	SSA_West	West	true	false	Female	55	26.2	Middle	Low	-1.093	false	II	3	5.33	Chemo_only	Luminal_A	Positive	Positive	Negative	Low	false	4.118	false
INTEGRATED_SAMPLE_00075	SSA_West	West	true	false	Female	52	27.8	Low	Low	-1.371	true	III	3	10.91	Chemo_plus_targeted	Luminal_A	Positive	Positive	Negative	Intermediate	false	6.29	true
INTEGRATED_SAMPLE_00076	SSA_West	West	true	false	Female	58	21.1	Middle	Average	-0.243	false	I	2	2.74	Chemo_plus_targeted	Luminal_A	Positive	Positive	Negative	High	false	5.19	false
INTEGRATED_SAMPLE_00077	SSA_West	West	true	false	Female	45	27.3	High	Low	-0.678	false	III	3	7.38	Chemo_plus_targeted	Basal_like	Negative	Negative	Negative	High	true	0.456	true
INTEGRATED_SAMPLE_00078	SSA_West	West	true	false	Female	52	27.8	Low	High	0.588	false	IV	2	6.18	Palliative_only	Luminal_A	Positive	Positive	Negative	High	false	0.026	true
INTEGRATED_SAMPLE_00079	SSA_West	West	true	false	Male	57	16.7	Low	High	1.235	false	I	3	0.8	Surgery_only	Basal_like	Negative	Negative	Negative	Low	true	0.172	true
INTEGRATED_SAMPLE_00080	SSA_West	West	true	false	Female	47	22.9	Low	Average	-0.644	false	IV	3	6.17	Chemo_only	Luminal_B	Positive	Negative	Negative	Low	false	0.777	true
INTEGRATED_SAMPLE_00081	SSA_West	West	true	false	Female	55	29.5	Low	Average	0.11	false	IV	3	5.04	Chemo_plus_targeted	Luminal_A	Positive	Positive	Negative	Low	false	0.974	true
INTEGRATED_SAMPLE_00082	SSA_West	West	true	false	Female	43	29.4	High	Average	0.309	false	II	1	4.98	Chemo_only	HER2_enriched	Positive	Positive	Positive	Intermediate	false	6.322	false
INTEGRATED_SAMPLE_00083	SSA_West	West	true	false	Female	46	21.7	Middle	Average	-1.183	false	III	3	6.74	Surgery_only	Luminal_A	Positive	Positive	Negative	Intermediate	false	1.451	true
INTEGRATED_SAMPLE_00084	SSA_West	West	true	false	Female	46	28.3	Low	Low	-0.932	false	II	3	4.94	Endocrine_only	Basal_like	Negative	Negative	Positive	Low	false	5.371	false
INTEGRATED_SAMPLE_00085	SSA_West	West	true	false	Female	37	32.6	Low	Average	-0.108	false	IV	3	7.86	Palliative_only	Luminal_A	Positive	Negative	Negative	Low	false	0.501	true
INTEGRATED_SAMPLE_00086	SSA_West	West	true	false	Female	55	31.4	Low	Low	-1.226	false	III	3	4.1	Chemo_only	Basal_like	Negative	Negative	Positive	Intermediate	false	2.506	false
INTEGRATED_SAMPLE_00087	SSA_West	West	true	false	Female	45	27	Low	Average	-0.272	false	IV	3	5.9	Palliative_only	Basal_like	Negative	Negative	Negative	Intermediate	true	0.202	true
INTEGRATED_SAMPLE_00088	SSA_West	West	true	false	Female	50	25.2	Middle	Average	-0.835	false	II	3	2.06	Chemo_only	Luminal_B	Positive	Positive	Negative	Low	false	2.431	false
INTEGRATED_SAMPLE_00089	SSA_West	West	true	false	Female	55	24.4	Middle	Average	-0.266	false	III	3	8.16	Chemo_only	Luminal_A	Positive	Positive	Negative	Intermediate	false	5.419	false
INTEGRATED_SAMPLE_00090	SSA_West	West	true	false	Female	55	29.8	Middle	Low	-0.84	false	III	2	3.93	Chemo_only	HER2_enriched	Positive	Positive	Positive	High	false	5.284	true
INTEGRATED_SAMPLE_00091	SSA_West	West	true	false	Female	57	28.5	Low	Average	0.104	false	II	2	5.4	Chemo_plus_targeted	Luminal_A	Positive	Positive	Negative	Low	false	4.315	true
INTEGRATED_SAMPLE_00092	SSA_West	West	true	false	Female	49	30.6	Middle	Low	-1.255	false	I	1	0.2	Palliative_only	HER2_enriched	Negative	Positive	Positive	High	false	2.896	false
INTEGRATED_SAMPLE_00093	SSA_West	West	true	false	Female	45	28.2	Middle	Average	-0.117	false	IV	3	9.49	Palliative_only	Basal_like	Negative	Negative	Positive	Low	false	0.106	true
INTEGRATED_SAMPLE_00094	SSA_West	West	true	false	Female	49	31.6	Low	Average	-0.026	false	III	3	4.77	Endocrine_only	Basal_like	Negative	Negative	Negative	Low	true	1.387	true
INTEGRATED_SAMPLE_00095	SSA_West	West	true	false	Female	31	28	Middle	High	0.134	false	II	3	4.08	Surgery_only	HER2_enriched	Positive	Negative	Positive	Low	false	3.401	true
INTEGRATED_SAMPLE_00096	SSA_West	West	true	false	Female	34	31.4	Middle	Average	-0.319	false	IV	3	8.94	Chemo_only	Basal_like	Negative	Negative	Negative	Intermediate	true	1.02	true
INTEGRATED_SAMPLE_00097	SSA_West	West	true	false	Female	35	31.9	Middle	Average	-0.49	false	II	3	0.52	Chemo_only	Luminal_A	Positive	Positive	Negative	Intermediate	false	6.913	false
INTEGRATED_SAMPLE_00098	SSA_West	West	true	false	Female	39	23.5	Low	Low	-1.161	false	III	3	8.86	Chemo_plus_targeted	Luminal_A	Positive	Positive	Negative	Intermediate	false	0.109	true
INTEGRATED_SAMPLE_00099	SSA_West	West	true	false	Female	54	18.5	Low	Low	-1.041	false	I	2	0.2	Surgery_only	Basal_like	Negative	Negative	Negative	Intermediate	true	2.528	true
INTEGRATED_SAMPLE_00100	SSA_West	West	true	false	Female	40	35.2	Middle	Low	-1.106	false	III	3	3.84	Chemo_only	Basal_like	Negative	Negative	Negative	Intermediate	true	1.841	true

End of preview. Expand in Data Studio

SSA Breast Integrated Genomic-Clinical ML Dataset (Multi-ancestry, Synthetic)

Dataset summary

This module provides an integrated, machine-learning–ready synthetic dataset that links:

Genotype (polygenic risk score, high-risk germline carrier flag).
Clinical and pathological phenotype (stage, grade, tumour size, ER/PR/HER2, subtype).
Key confounders (age, BMI, socioeconomic status, treatment regimen).
Longitudinal biomarkers (tumour burden and neutrophil-to-lymphocyte ratio trajectories).
Time-to-event outcomes (overall survival time and event indicator).

The design is grounded in:

TCGA-BRCA and METABRIC (integrated DNA/RNA/clinical + survival).
PRS studies in African ancestry women (attenuated AUC; modest risk gradients).
Registry/meta-analytic survival literature (stage-specific OS, SES effects, BMI effects).
Methodological work on joint longitudinal–survival models (dynamic risk from NLR/tumour burden trajectories).

All observations are fully synthetic, with distributions tuned qualitatively to reflect published patterns while avoiding re-identification or direct reproduction of any individual-level dataset.

Cohort design

Sample size and populations

Total N: 10,000 synthetic invasive breast cancer cases.
Populations:
- SSA_West: 2,000
- SSA_East: 2,000
- SSA_Central: 1,500
- SSA_Southern: 1,500
- AAW (African American women): 1,500
- EUR (European reference): 1,000
- EAS (East Asian reference): 500
Sex:
- Predominantly Female (~99%), with ~1% male breast cancers.
Age:
- 18–90 years.
- Slightly younger mean ages in SSA cohorts; older in AAW/EUR/EAS, reflecting registry patterns.
Follow-up:
- Intended follow-up window ∈ [0.5, 10] years.
- Survival times are simulated from a parametric hazard model with multiplicative effects of genotype, stage, subtype, treatment, BMI, SES, and NLR.

Confounders and clinical context

Each patient has:

age (years) – population-specific normal distributions, truncated to 18–90.
bmi (kg/m²) – continuous; higher means in SSA_Southern and AAW, lower in EAS.
ses – categorical: Low, Middle, High; distributions vary by population, with higher Low fractions in SSA and AAW compared with EUR/EAS.
stage_at_diagnosis – I, II, III, IV; more advanced stages in SSA/AAW relative to EUR/EAS, tied to literature on late presentation.
grade – 1/2/3; conditional on stage, with higher emphasis on grade 3 at stages III–IV.
tumour_size_cm – continuous; stage-dependent means, mimicking clinical T-stage.
treatment_regimen – Surgery_only, Endocrine_only, Chemo_only, Chemo_plus_targeted, Palliative_only; distributions conditioned on stage, with more Palliative_only and less curative surgery in stage IV.

These variables serve as confounders and mediators for genotype–outcome relationships in downstream ML models.

Genotype representation

Two genotype-related constructs are simulated:

Polygenic Risk Score (PRS):
- prs_category – Low, Average, High with population-specific distributions that allow somewhat higher fractions of High in AAW and some SSA groups.
- prs_standardized – continuous standardized PRS (mean/SD per category) with moderate effect on hazard.
- Design is informed by African-ancestry PRS studies (e.g. Yao et al., Gao et al.), which show risk stratification but attenuated AUC compared with EUR-derived scores.
High-risk germline carrier flag:
- high_risk_germline – boolean approximating rare pathogenic variants (e.g., BRCA1/2, TP53), with prevalence 2–3.5% depending on population.

In the survival model, both PRS and high-risk carrier status contribute modestly to log-hazard, avoiding unrealistic effect sizes.

Phenotypes and tumour biology

Molecular subtype and receptors

Variables:

tumor_subtype – Luminal_A, Luminal_B, HER2_enriched, Basal_like, Normal_like.
er_status – Positive / Negative.
pr_status – Positive / Negative.
her2_status – Positive / Negative.
triple_negative – derived boolean (ER-/PR-/HER2-).

Anchoring:

Subtype distributions by population mirror PAM50-like patterns:
- Higher Basal_like in SSA_* and AAW.
- Higher Luminal_A in EUR/EAS.
ER/PR/HER2 probabilities are subtype-dependent, enforcing concordance (e.g., almost all Luminal_A are ER+/PR+, most HER2_enriched are HER2+).

Inflammation and NLR

Variables:

nlr_category – Low, Intermediate, High at baseline.
Longitudinal nlr_value trajectories in the time-series table (see below).

Anchoring:

Based on systemic immune marker meta-analyses, High NLR is more prevalent in SSA/AAW compared with EUR/EAS.
High NLR contributes modestly to higher hazard in the survival model.

Survival model

The survival time (os_time_years) and event indicator (os_event) are simulated from a parametric proportional hazards–style model:

Baseline hazard per year (~0.12) with log-normal frailty.
Additive log-hazard contributions from:
- Age (per 10 years above ~50).
- BMI (per 5 kg/m² above ~27).
- Standardized PRS.
- High-risk germline flag.
- Stage at diagnosis (strong gradient from I to IV, consistent with meta-analytic estimates).
- Tumour subtype (worse hazard in Basal_like and HER2_enriched).
- SES (higher hazard for Low vs High).
- Treatment regimen (better outcomes with curative surgery/systemic therapy, worse with Palliative_only).
- NLR category (higher hazard with High vs Low).

Censoring times are drawn uniformly between 0.5 and 10 years, ensuring a mix of events and censored observations.

Validation enforces that mean OS decreases from stage I → II → III → IV, and that High PRS has worse mean survival than Low PRS by a modest margin.

Longitudinal time-series

A separate time-series table records repeated measurements at preset time points:

time_months: 0, 6, 12, 24.
tumour_burden_cm: approximate lesion size / composite tumour burden.
nlr_value: continuous NLR-like value.
response_class: latent treatment response pattern.

Response classes

Good_response: rapid burden reduction and mild NLR decline.
Intermediate_response: modest initial reduction or plateau.
Poor_response: increasing burden and rising NLR.

Response class is sampled conditional on baseline stage, with more Poor_response in stage IV.

Tumour burden and NLR trajectories apply class-specific multipliers and means from the configuration, with small log-normal/normal noise to preserve variability.

Validation checks ensure that:

Good_response has clearly lower mean tumour burden and NLR at 24 months vs baseline.
Poor_response shows increasing tumour burden and NLR over time.

Files and schema

Baseline table

Files:

integrated_ml_baseline.parquet
integrated_ml_baseline.csv

Each row is one patient with:

Demographics & confounders:
- sample_id
- population, region, is_SSA, is_reference_panel
- sex
- age
- bmi
- ses
Genotype:
- prs_category
- prs_standardized
- high_risk_germline
Clinical & tumour:
- stage_at_diagnosis
- grade
- tumour_size_cm
- treatment_regimen
Phenotypes:
- tumor_subtype
- er_status, pr_status, her2_status
- triple_negative
- nlr_category
Survival:
- os_time_years
- os_event

Time-series table

Files:

integrated_ml_timeseries.parquet
integrated_ml_timeseries.csv

Each row is one measurement time for a patient:

sample_id
population
stage_at_diagnosis
response_class
time_months
tumour_burden_cm
nlr_value

Generation

The dataset is generated using:

integrated_ml/scripts/generate_integrated_ml.py

with configuration in:

integrated_ml/configs/integrated_ml_config.yaml

and literature inventory in:

integrated_ml/docs/LITERATURE_INVENTORY.csv

Key generation steps:

Demographic and confounder sampling – multi-ancestry population, age, BMI, SES.
Genotype sampling – PRS category + standardized PRS and a rare high-risk germline flag with population-specific prevalences.
Clinical and phenotype assignment – stage, grade, tumour size, treatment regimen, subtype, receptor status, triple-negativity, baseline NLR category.
Survival time simulation – log-hazard model combining genotype, phenotype, confounders, and baseline NLR, plus random frailty and censoring.
Response-class and trajectory generation – latent response class per patient and longitudinal tumour burden/NLR trajectories at fixed timepoints.

Validation

Validation is performed with:

integrated_ml/scripts/validate_integrated_ml.py

and summarized in:

integrated_ml/output/validation_report.md

Checks include:

C01–C02 – Sample size and population counts vs configuration.
C03 – Stage distributions by population.
C04 – SES distributions by population.
C05 – Tumour subtype distributions by population.
C06 – Survival gradient by stage (I > II > III > IV, with ≥3 years difference in mean OS between I and IV).
C07 – PRS effect on survival (worse mean OS in High vs Low PRS).
C08 – Time-series response patterns (Good vs Poor response classes).
C09 – Missingness across baseline and time-series key variables.

Intended use

This integrated dataset is intended for:

End-to-end ML experiments where genotype, clinical, and longitudinal data feed into survival prediction.
Fairness and confounding analysis, e.g.:
- Assessing how SES, BMI, and treatment mediate survival differences.
- Exploring PRS calibration across ancestry groups.
Method development for joint models that incorporate repeated biomarkers and time-to-event endpoints.
Educational use to teach students:
- How real-world EHR/registry data might be structured.
- How confounding, mediation, and selection bias can appear.

It is not intended for:

Estimating real-world effect sizes or survival rates.
Deriving clinical practice guidelines, risk thresholds, or patient-level clinical decisions.

Ethical considerations

All records are synthetic, generated from parameterised models informed by published group-level statistics.
Ancestry and SES patterns are modeled to emulate known disparities; they must not be used to stigmatise or essentialise individuals or groups.
Any risk or fairness analyses performed on this dataset should be contextualised within broader socio-economic and structural determinants of health.

License

License: CC BY-NC 4.0.
Free for non-commercial research, method development, and education with attribution.

Citation

If you use this dataset, please cite:

Electric Sheep Africa. "SSA Breast Integrated Genomic-Clinical ML Dataset (Multi-ancestry, Synthetic)." Hugging Face Datasets.

and, where applicable, foundational resources such as TCGA, METABRIC, African-ancestry PRS studies, and methodological work on longitudinal–survival modeling.

Downloads last month: 22

Size of downloaded dataset files:

451 kB

Size of the auto-converted Parquet files:

451 kB

Number of rows:

50,000

Collection including electricsheepafrica/ssa-breast-integrated-ml

Genomics

Collection

A collection of datasets on genomics in Sub-Saharan Africa • 21 items • Updated 6 days ago