Computer assisted substance identification in systematic toxicological analysis: New life for old methods?

Abstract 

Systematic toxicological analysis is a major task in clinical and forensic toxicology, usually being addressed with mass spectrometric methods. As an alternative to approach substance identification in resource limited laboratories, we developed an Internet-based software to process analytical data, obtaining corrected identification parameters and using a probabilistic approach to retrieve a database comprised of retention indices in gas chromatography and relative retention times and absorption maxima in the UV range in high performance liquid chromatography, obtained at standardized conditions. The probabilistic approach allowed ranking candidate substances on basis of a Similarity Index. A Mean List Length of 1.6 was obtained when all analytical data was informed, considering a database of 1762 compounds. In 34 of 35 spiked samples, the software ranked the correct substance at the higher Similarity Index. List lengths similar to those obtained with more sophisticated methods can be found with the software, which represents a valuable alternative to resource limited laboratories.

Keywords: Systematic toxicological analysis, Substance identification, Database retrieval, Chromatographic analysis, Software development