Figure 2.

Quantitative evaluation of LC/MS profile alignment using apLCMS. Top row shows Pearson correlation within sample duplicates in both datasets; bottom row shows the median PID of feature intensities within sample duplicates. The effect of re-aligning profiles after removing poor quality samples (correlation coefficient, R2 < 0.7) on the quantitative reproducibility of features is shown in the bottom right panel. A noticeable difference in median PID can be seen between alignment using all samples and alignment using only high quality samples for both columns of Sample Set 2.

Uppal et al. BMC Bioinformatics 2013 14:15   doi:10.1186/1471-2105-14-15
Download authors' original image