From my background reading it is suggested you use internal standards in each LC-MS sample in order to improve the quality of data analysis. However, one of the examples given is of ~8 isotopic compounds such as L-lysine d4 of which many are very expensive and others do not mention what the internal standards are. Is there a minimum number of standards required for a robust experiment or are they specifically dependent on the experiment itself? I currently use pooled QC samples and use reference compounds to calibrate the machine but nothing else.