Retrieval of analytical data and substance identification in systematic toxicological analysis by the mean list length approach.

Chromatographic techniques are basic tools in systematic toxicological analysis. Extensive data bases with retention parameters of known drugs to aid in the identification of substances found are available or in preparation. For a search in such a data base the computer is indispenable. The commonly used window search has some disadvantages which can be overcome by a search based on the statistical concept, the mean list length. The latter retrieval system gives for each candidate in the identification process a probability value. It is shown that these probability values are highly influenced by the reproducibility of the retention parameters of the analytical systems used. Explanations for these phenomena are given.