Development of an automated data processing method for sample to sample comparison of seized methamphetamines

The information about the sources of supply, trafficking routes, distribution patterns and conspiracy links can be obtained from methamphetamine profiling. The precursor and synthetic method for the clandestine manufacture can be estimated from the analysis of minor impurities contained in methamphetamine. Also, the similarity between samples can be evaluated using the peaks that appear in chromatograms. In South Korea, methamphetamine was the most popular drug but the total seized amount of methamphetamine whole through the country was very small. Therefore, it would be more important to find the links between samples than the other uses of methamphetamine profiling. Many Asian countries including Japan and South Korea have been using the method developed by National Research Institute of Police Science of Japan. The method used gas chromatography-flame ionization detector (GC-FID), DB-5 column and four internal standards. It was developed to increase the amount of impurities and minimize the amount of methamphetamine. After GC-FID analysis, the raw data have to be processed. The data processing steps are very complex and require a lot of time and effort. In this study, Microsoft Visual Basic Application (VBA) modules were developed to handle these data processing steps. This module collected the results from the data into an Excel file and then corrected the retention time shift and response deviation generated from the sample preparation and instruments analysis. The developed modules were tested for their performance using 10 samples from 5 different cases. The processed results were analyzed with Pearson correlation coefficient for similarity assessment and the correlation coefficient of the two samples from the same case was more than 0.99. When the modules were applied to 131 seized methamphetamine samples, four samples from two different cases were found to have the common origin and the chromatograms of the four samples were appeared visually identical. The developed VBA modules could process raw data of GC-FID very quickly and easily. Also, they could assess the similarity between samples by peak pattern recognition using whole peaks without spectral identification of each peak that appeared in the chromatogram. The results collectively suggest that the modules would be useful tools to augment similarity assessment between seized methamphetamine samples.