A set of data on a given peroxidase inhibition by quecertin, showing a
n unusually high experimental error, was used to demonstrate how data,
seemingly unsuitable for graphical analysis, may still provide useful
information on the inhibition mechanism. The most reliable model turn
ed out to be a mixed non-competitive inhibition. The present statistic
al procedure has proved (i) to be a simple, general and unequivocal wa
y to carry on kinetic analysis of enzymatic reactions under inhibition
action, based only on consecutive linear regressions (no previous ass
umption about the inhibition mechanism is required); (ii) that experim
ental errors can play an essential role on the final decision about th
e inhibition mechanism and (iii) to be able to show how close to the M
ichaelis-Menten mechanism the kinetic model actually is. Therefore all
proposed inhibition mechanisms were subjected to statistical judgemen
t.