Identification and rejection of outliers in enzyme kinetics.

A program (AICOUT) for the correct choice of the experimental value and weight for replicate enzyme kinetic determinations is given. It is based on the method of identification of outliers proposed by Kitagawa (Technometrics, 21 (1979) 193-199). The program is written in BASIC and FORTRAN77. The FORTRAN77 version of AICOUT program coupled to a FORTRAN77 version of the non-linear regression program previously published by Canela (Int J Biomed Comput, 15 (1984) 121-130) is given. This joint program leads to an improvement of precision and confidence in the estimated parameters when the suitable strategy is used. This strategy is as follows: (i) the experimental points are selected, (ii) several replicates of each point are performed, (iii) data are analyzed and outliers are rejected, (iv) normal or biweighted regression is carried out.