Fig. 2.

Background error estimation workflow. (A) The first step runs over every position in all control samples and counts the total occurrences of every A, C, G and T. It also stores the average base quality score for each position. (B) The second step’s goal is to remove any suspected variant from the pileup as our objective is to estimate background error noise only. (C) In this step, the counts are converted to probabilities by dividing them by the depth for each position. (D) The final step consists of converting the base quality score of each position to the corresponding ASCII + 33 character.