Allow for filtering of loaded metabolomics data

Add the option to parse a user-supplied list of strain labels for a given dataset (in addition to the regular strain_mappings.csv file), and during loading optionally discard any metabolomics objects which don't have at least? any? strains in the list