Inaccuracies in the quantification of peptides - a case study using beta-endorphin assay