Estimating Phred scores of Illumina base calls by logistic regression and sparse modeling

BMC Bioinformatics

Table 3 Four combinations of schemes between two base-calling and two quality scoring methods

		Quality scoring
		L1-regularized logistic regression	Lookup table strategy
Base-calling	Bustard	Bustard+Logistic	Bustard+Lookup
	3Dec	3Dec+Logistic	3Dec+Lookup

We have two base calling methods: Bustard, the default method embedded in the Illumina sequencers; 3Dec, our newly developed method. We also have two quality scoring methods: Lookup, the lookup table strategy adopted by Illumina; Logistic, the L ₁-regularized logistic regression model proposed in this study

ISSN: 1471-2105