2.6 Data Processing
Waters raw data files generated from the metabolomics analysis were
converted to mzData files using the convert.waters.raw R package. Using
the isotopologue parameter optimization (IPO) R package, the quality
control pooled sample injections were used to find the optimal peak
processing parameters, retention time corrections and grouping
parameters. Parameters generated from IPO were used for XCMS processing
of metabolomics data. The CAMERA package was applied to XCMS processed
features to annotate possible isotopes and adducts. The resulting data
was subsequently normalized to internal standards, and features with
>30% relative standard deviation (rsd) within quality
control injections were excluded from analysis. Urine features were
normalized to their corresponding urinary creatinine and internal
standard signals, to account for differences in urine concentration.
Using the CAMERA package, features were grouped based on Pearson
correlation coefficients and retention time into “pcgroups”. Within
each pcgroup, only the feature with the highest mean raw intensity was
kept for further data analysis. Duplicate features found in both the
untargeted and derivatized experiments were removed from the derivatized
dataset before analysis. The raw intensity values of all features were
log transformed using MetaboAnalyst 5.0, to remove heteroscedasticity
and correct for skewed data distribution. Any 0 values during log
transformation were treated as 1/5 of the minimum intensity values of
each feature. Log transformed feature intensity values were used for all
analyses unless stated otherwise.