The validity of comparing the measurements of angiotensin I generated in human plasma by radioimmunoassay and bioassay.