Dataset: MS metabolomics of blue crab urine

MS data were aligned with an adaptive curve and metabolic features (retention time (Rt), m/z pairs) were detected after removal of isotopes and adducts in Compound Discoverer. Background peaks were removed from the dataset using a sample blank. Fill gaps and normalization by constant sum was applied to the data. Select features with a fold change greater or equal to 2 in urine from blue crabs fed mud crabs when compared to urine from blue crabs fed oysters and relative standard deviation of less than 30 were used to build an un-supervised PCA model with leave-one-out cross-validation using MATLAB R2012b (Version 8.0.0.783 The MathWorks, Inc.) and the PLS Toolbox (v.6.71, Eigenvector Research, Inc.). Data were preprocessed by auto-scaling the features’ peak areas across the samples.