出版社：Japanese Association of Forensic Science and Technology
摘要：One of the changes in the version update to v1.6 of GeneMapper ID-X Software (GMIDX) designed for human identification data analysis is a revision of the peak height detection algorithm. Hence, the evaluation of impact to the data interpretation of STR typing by the differences of the software version is required to utilize the latest version for routine casework. In this study, concordance study for calculated peak details and quality assessment by process quality value (PQV) flags was performed by comparing data for a total of 1844 electrophoresis run data including controls and allelic ladder analyzed using GMIDX v1.4 and v1.6, respectively. For peak details, while sizing and data point of the peaks analyzed using each version were all concordant for the peaks detected in both versions, discordance was found in allele call, peak height, and peak area. For discordance of peak height and peak area, the values calculated by GMIDX v1.6 were all smaller than those by GMIDX v1.4. Some allelic dropouts which the peak was detected by one version was not detected by the other version. For the quality assessments by PQV flags, discordance was found in only SPK in Sample-Level Quality Assessments, discordance was found in BIN, PHR, LPH, AN, BD, SPK, and ACC in Marker-Level Quality Assessments. These discordances except for peak area were caused by lower peaks derived from artifact and/or noise. Further, the impact of the version difference on the calculation of the minimum thresholds was evaluated. Although slight differences in detected peak height were observed by the revision of the peak height detection algorithm, the calculated minimum thresholds resulted in the same values.
