Impact of pre-imputation SNP-filtering on genotype imputation results

BMC Genomic Data

Table 3 (Imputation quality of the scenario “Hole-filling with external HapMap reference”): percentages of overlapping masked genotypes imputed with good Hellinger score (≥0.6) are presented

Datasets		MACH Imputation score based on 10% overlapping masked genotypes			IMPUTE2 Imputation score based on 10% overlapping masked genotypes
Data subset name	#SNPs	10%	20%	50%	10%	20%	50%
ALL	9602	**94.03+***	93.44*	**91.03+***	*93.12+*	*92.42+*	*89.74*
LQ	8472	94.01*	**93.45+***	91.01*	*93.10*	*92.41*	*89.85+*
NQ	7923	93.83*	93.15*	90.61*	92.90	92.11	89.32
BQ	6337	90.85*	89.62*	83.34*	89.85	88.61	82.47
HQ	4658	90.47*	89.05*	81.84*	89.00	87.55	80.57

Datasets of different pre-imputation quality filtering were considered and different percentages of genotypes were masked. Results of the optimal imputation scenarios are described with (⁺). Results of the filtering scenarios which are not significantly inferior compared to the best scenario, are described with Italic-bold letters. An asterisk (*) indicates whether MaCH or IMPUTE2 performed significantly better in the corresponding scenario.

ISSN: 2730-6844