ReQON: a Bioconductor package for recalibrating quality scores from next-generation sequencing data

Abstract

Background: Next-generation sequencing technologies have become important tools for genome-wide studies. However, the quality scores that are assigned to each base have been shown to be inaccurate. If the quality scores are used in downstream analyses, these inaccuracies can have a significant impact on the results.

Results: Here we present ReQON, a tool that recalibrates the base quality scores from an input BAM file of aligned sequencing data using logistic regression. ReQON also generates diagnostic plots showing the effectiveness of the recalibration. We show that ReQON produces quality scores that are both more accurate, in the sense that they more closely correspond to the probability of a sequencing error, and do a better job of discriminating between sequencing errors and non-errors than the original quality scores. We also compare ReQON to other available recalibration tools and show that ReQON is less biased and performs favorably in terms of quality score accuracy.

Conclusion: ReQON is an open source software package, written in R and available through Bioconductor, for recalibrating base quality scores for next-generation sequencing data. ReQON produces a new BAM file with more accurate quality scores, which can improve the results of downstream analysis, and produces several diagnostic plots showing the effectiveness of the recalibration.

Figure 1

Recalibration of U87 cell line replicate 1 with ReQON. Plot A shows the distribution of sequencing errors by read position. Plot B shows frequency distributions of quality scores before (solid blue) and after (dashed red) recalibration. Reported quality scores versus empirical quality scores are shown before (plot C) and after (plot D) recalibration. The points are shaded according to the frequency of bases assigned that quality score, corresponding to the values shown in plot B. Plots C and D also report the Frequency-Weighted Squared Error (FWSE), a measure of quality score accuracy. The large decrease in FWSE confirms that the recalibrated quality scores more accurately represent the probability of a sequencing error than the original quality scores.

Figure 2

Discrimination performance of original and ReQON-recalibrated quality scores. Relative frequency distributions of quality scores for bases not matching the reference sequence in chromosome 20 of cell line replicate 2. These non-reference bases are separated as belonging to positions in dbSNP version 132 (known variants, red curve) versus other positions (sequencing errors, blue curve). Plot A shows the distribution of original quality scores and plot B shows the distribution after recalibration with ReQON. The area under the ROC curve (AUC) is reported. The increased AUC demonstrates that the recalibrated quality scores do a better job of distinguishing sequencing errors from non-errors.

Figure 3

Example position where bases are identified as sequencing errors by GATK but not ReQON. Plot A shows an Integrative Genomics Viewer (IGV) visualization of chr10:75,531,679-75,531,712 for cell line replicate 1, highlighting a position where the reference sequence is T but all of the bases mapped to this position are a C. This position (chr10:75,531,700) is not listed as a known variant in dbSNP version 132. The bases at this position are removed from the training set by ReQON but are called as sequencing errors by GATK. Plot B shows box plots comparing the quality scores of the bases at this position after recalibration with GATK and ReQON. Overall, ReQON assigns higher quality scores to these non-reference bases than GATK.