Figure 2

Comparison of the two-stage preprocessing approach with traditional apLCMS and XCMS using standard sample. Each dot represents a parameter setting. (a) Total number of zeros in the final data matrix; (b) proportion of features with m/z matched to known metabolites using xMSAnnotator; (c) level of variation as measured by coefficient of variation (CV) in the final data matrix without considering batches; (d) level of variation as measured by coefficient of variation (CV) in the final data matrix without considering batches, considering only non-zero values; (e) level of variation as measured by CV after merging each batch; (f) level of variation as measured by CV after merging each batch, considering only non-zero values. In all CV plots, the point is median; vertical bars represent 10th to 90th percentile.