. 2018 Sep 14;9:3753. doi: 10.1038/s41467-018-05936-5

Table 1.

Goodness-of-fit statistics for mutation rate estimates applied to de novo testing data

Mutation rate estimation strategy			AIC	ΔAIC^a	AIC rank^b	Nagelkerke’s R²
Subtype length	Study	Variant type	AIC	ΔAIC^a	AIC rank^b	Nagelkerke’s R²
1-mers	BRIDGES	ERVs	353,896	21,575	7	0.088
3-mers	BRIDGES	ERVs	335,319	2998	4	0.118
5-mers	BRIDGES	ERVs	332,861	540	3	0.124
7-mers	BRIDGES	ERVs	332,321	0	1	0.126
7-mers	BRIDGES	ERVs (passing 1000G strict mask)	332,582	261	2	0.125
7-mers	BRIDGES	MAC10+	342,886	10,565	5	0.103
7-mers	1000G	Intergenic SNVs⁷	344,003	11,682	6	0.100

^aDifference in AIC from the baseline BRIDGES 7-mer model

^bLower AIC rank indicates better model performance