AccuPRO-ID® Performance Evaluation: A Case Study
Accurate classification of unknown bacterial isolates is an essential first step in understanding the impact organisms have on an environmental monitoring program. Accugenix offers the AccuPRO-ID® solution as an ideal technology for accurate, reproducible, fast and inexpensive microbial identification for industries required to identify microorganisms on a regular basis. This solution is a proteotypic service that utilizes matrix-assisted laser desorption/ionization-time of flight (MALDI-TOF) mass spectrometry and, since this is a technology in its infancy for commercial bacterial identifications, we support the technology with our proprietary genotypic 16S rDNA sequencing technology, AccuGENX-ID™.
The library databases at Accugenix are the most extensive of all the commercial libraries available. In fact, the libraries from all commercially available systems, both genotypic and phenotypic, are merely a subset of the Accugenix reference libraries. See Library Comparison for more information. It is critical that the identification libraries, against which you compare your data whether generated by phenotypic, proteotypic or genotypic methods, contain all species relevant to your environment. If the library lacks depth of coverage, the interpretation of your data may not be reliable.
In combination with maintaining relevant libraries, having consistent qualified methods of analysis is essential when identifying organisms in the manufacturing environment. MALDI-TOF technology is highly accurate and less dependent on growth conditions than phenotypic methods, because it analyzes primarily ribosomal proteins which are constitutively expressed at very high levels.
In one case study presented below, Accugenix evaluated the accuracy of reported identifications from the AccuPRO-ID® system to the results from the bioMérieux VITEK® 2. The data show that the AccuPRO-ID® service from Accugenix is significantly more accurate than the VITEK® 2, leading to more confidence in the Environmental Monitoring information and allowing for more effective trending and tracking in a production facility.
Microbial identification at Accugenix. All samples were processed according to the AccuPRO-ID® solution and were subjected to MALDI-TOF analysis. MALDI-TOF yielded results for the majority of the samples. Under the usual AccuPRO-ID® offering, the samples that did not yield significant spectra would have been automatically tested with 16S rDNA sequencing (refluxed to sequencing). However, as part of this blinded comparative study, all samples were subjected to 16S rDNA sequencing as it served as the reference method to verify the taxonomic identity of the organisms. The microbial identifications generated by the VITEK® 2 (performed by our client), Accugenix’s MALDI-TOF and Accugenix’s 16S rDNA sequencing were compared to determine the accuracy of the phenotypic and proteotypic methods, respectively, the results of which are shown in the Match to Sequence Reference columns in Table 1.
As compared to the AccuGENX-ID™ sequence-based identification, the VITEK® 2 was correct to the species level for 41% of the samples, correct to the genus level 15% of the time and was incorrect on 44% of the identifications (Figure 1). Interestingly, the average Match Factor in each of these groups was very similar (Figure 1). Of the 44% that were completely misidentified, the mean confidence level was 94%. In contrast to the performance of VITEK® 2, the AccuPRO-ID® method described here was correct to the species level on 90% of the samples and correct to the genus level with the remaining 10% of the samples (Figure 2). As we typically see with the MALDI-TOF BioTyper, approximately 31% of the samples are sent to sequencing (Figure 2). Overall, these results suggest that the AccuPRO-ID® method is more accurate than the VITEK® 2.
Further analysis of the 17 samples incorrectly identified by VITEK® 2 (Table 1) showed that only three of the correct entries are present in the VITEK® 2 database (C. tuberculostearicum, C. aurimucosum, C. mucifaceiens). Additionally, 3 individual samples were all identified as C. tuberculostearicum by AccuPRO-ID®, however, the VITEK® 2 generated three different IDs for the same organism (Kocuria rosea, Kytococcus sedentarius and Micrococcus luteus). Having three different identifications for the same organism makes accurate trending and tracking nearly impossible to accomplish with the VITEK® 2 considering the identification of the same organism can be so highly variable.
Conclusions. Current available methods of identification range from genotypic to phenotypic, with 16S sequencing being the gold standard for bacterial identification. Accugenix provides reference quality genotypic methods for identification and has maintained a bacterial library focusing on organisms relevant to all industries we serve. When identifications are based on phenotypic characteristics, such as with the VITEK® 2 Compact, the methods are highly error prone, variable and subjective. With the introduction of the proteotypic MALDI-TOF-based method of identification that is supported by sequencing, the AccuPRO-ID® solution, Accugenix is providing an additional highly accurate, fast and inexpensive option for routine monitoring programs. See AccuPRO-ID vs. VITEK Case Studies for an additional comparative study.
