Explicable prioritization of genetic variants by integration of rule-based and machine learning algorithms for diagnosis of rare Mendelian disorders

Table 2 Top-k recall rates for different 3ASC models compared with baseline

Models	Top 1	Top 3	Top 5	Top 10	Top 20	Top 30	Top 100
3ASC (Baseline)	0.518	0.742	0.825	0.911	0.969	0.986	1.000
3ASC_LR	0.591	0.785	0.862	0.937	0.977	0.991	1.000
3ASC_RF	0.718	0.841	0.889	0.936	0.953	0.960	0.992
3ASC_LR_QC	0.718	0.847	0.915	0.966	0.989	0.995	1.000
3ASC_RF_QC	0.825	0.923	0.955	0.976	0.983	0.985	0.998
3ASC_LR_ALL	0.739	0.861	0.924	0.975	0.993	0.997	1.000
3ASC_RF_ALL	0.856	0.944	0.967	0.981	0.985	0.988	0.998

ISSN: 1479-7364