A statistical technique based upon a hyperbolic model has been developed for calibrating the standard dose-response cuvre of radioimmunoassay data, with subsequent calculation of unknown sample concentrations and their 95% confidence limits. A computer program based on this hyperbolic model has been written and is described here.In recent years, several aspects of mathematical and statistical modeling and analysis of radioimmunoassay (RIA) data have been published elsewhere (1-8). Our initial interest in the statistical problem of RIA analysis was initiated by the work of Bliss and Taljedal (1-4). We have since then developed a time-sharing FORTRAN program for the analysis of RIA data based upon a hyperbolic model whose parameters are estimated by the weighted least squares method. Some consideration of assay accuracy, precision, and sensitivity has been incorporated into this program. It is designed for general use in radioassays of steroids and proteins and for general drug screening in which the radioactivity is counted, regardless of the isotope employed. This work was done at G. D. Searle and Company, Chicago, Illinois, 60680. Wen-dar Cheng's present address is: The Dow Chemical Company, P. 0. Box 6851 I , Indianapolis, Indiona 46268.The total analysis program consists of three portions: assay control information, calculation and a summary table. The assay control information describes all sample tubes in detail (sequence and heading information). The calculation includes fitting of the standard curve, analysis of variance, estimation of assay error and parameters, and evaluation of unknowns. The third portion produces both tabular and graphical summaries of results. Figure 1 shows the general sequence of this RIA analysis program.This analysis program has been operating on the Honeywell 6000 computer (CECOS Operating System) for over a year, with a storage requirement of less than 7K words of core. The major advantages of this program include i t s simplicity, with stepwise and interactive operation, its statistical treatment of assay data, and its tendency to facilitate interassay comparison with respect to t h e parameters estimated. A portion of the analysis program described here has been incorporated into a complete on-line radioassay system developed by Searle Analytic, Inc. This on-line system automates the entire r a d i o a s s a y p r o c e d u r e , f r o m a n a l y z i n g radioimmunoassay/compe ti tive protein binding samples to prin t-out of immediate, meaningful results.It has been assumed that the random error follows a normal distribution with a variance 02 which is estimated as the within assay variance derived from the analysis of variance. This estimate of 02 is incorporated in the DRUG INFORMATION JOURNAL JanuarylJune 74 P7
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.