APHL 2019 POSTER ABSTRACTS
Use of Galileo™ AMR for the Automated Annotation of
Antimicrobial and Antiviral Resistance Genes in GenBank Evaluation of Leading Commercial Assays for the Early
Diagnosis of Lyme Disease in US Populations
G. Tsafnat, A. Watts and T. Watson, Arc Bio LLC J. Hahm, J. Breneman, J. Liu, L. Rabkina, W. Zheng, S. Zhou, R.
Walker and R. Kaul, Bio-Rad laboratories
Introduction: Galileo™ AMR is a cutting-edge software system for the
comprehensive annotation of antimicrobial resistance (AMR) genes
and associated elements that allow their mobilization. Galileo™
AMR supersedes the Multiple Antimicrobial Resistance Annotator
[1] developed at the New South Wales Univ and Westmead Hosp,
Sydney Australia and adds important features like high-throughput
(HT) analysis.
Galileo™ AMRlets users upload bacterial sequences. In a few
minutes, Galileo™ AMR accurately annotates all known Gram-
negative resistance genes and associated mobile elements,
including resistance gene cassettes, transposons etc. The High-
Throughput (HT) version of Galileo AMR™ can additionally annotate
SCCMec regions in Staphylococcus aureus and antiviral resistance
(AVR) markers in human cytomegalovirus (CMV).
Results: Over 99% of the elements in the Galileo™ AMR database
were found at least once in GenBank.
Gram-negatives: Galileo™ AMR found over 17,282 sequences
containing resistance genes and/or associated mobile elements.
Approximately half the sequences (n=8,948;51.8%) contained an
insertion sequence and over a quarter (n=4,737;27.4%) included
a transposon. Resistance genes were found for all classes of
antibiotics with beta-lactamase genes being the most common.
S. aureus: Approximately 1.6 million S. aureus sequences were
identified. Less than 6% showed evidence of the presence of
SCCMec.
CMV: All CMV sequences of containing the genes UL54 (n=399) and
UL97 (n=630) were annotated. All genes included at least one AVR
markers (range [1:56] each gene).
Conclusion: GenBank provided a rich catalogue of AMR sequences
in the studied organisms. Although the frequencies of exemplars
do not reflect frequencies in the real world, the database
represents drug resistance in the studied organisms. These results
demonstrate the ability of Galileo™ AMR to detect antimicrobial
resistance markers. Powered by multiple state-of-the-art databases
and a next generation genomic detection engine, Galileo™ AMR is
designed to empower those who perform AMR sequence research
and analysis.
Disclaimer: Galileo™ is a trademark of Arc Bio, LLC. For Research Use Only. Galileo™
AMR HT is currently under development
References: 1. Partridge SR, Tsafnat G, et al. FEMS Microbiol Rev. 2009
Jul;33(4):757-84.
Presenter: Guy Tsafnat, Arc Bio LLC, Menlo Park, CA
[email protected]
70
LAB MATTERS Summer 2019
Methods: An externally-validated Lyme ‘Premarketing’ panel from
the Centers for Disease Control consisting of 280 well-characterized
Lyme disease and non-disease controls was used to evaluate the
sensitivity and specificity of the BioPlex 2200 Lyme Total Assay. In
addition, 105 vendor provided Lyme disease positive samples and
200 endemic and non-endemic hospital normals were evaluated
during the course of this investigation. Results obtained were
compared to LIAISON Borrelia burgdorferi (DiaSorin, Inc.); Vidas II
IgM and IgG (BioMerieux, Inc.); Lyme PepC10/VlsE1 IgM/IgG (Zeus,
Inc.); Lyme C6 B. burgdorferi ELISA (Immunetics, Inc.). Positive or
equivocal samples were subsequently tested by both IgM and IgG
Western blot (MarDx, Inc.) assay.
Results: Using the CDC Lyme Premarketing panel, overall clinical
sensitivities of 91.1% (82/90), 84.4% (76/90), 88.9% (80/90),
and 85.6% (77/90) with specificities of 96.8% (184/190), 97.9%
(186/190), 89.5% (170/190), and 84.7% (161/190) were
respectively measured for the BioPlex 2200 Lyme Total assay,
Immunetics C6, Zeus PepC10/VlsE1, and BioMerieux VIDAS Lyme
II. Notably, sensitivities of 84.6% (33/39), 71.8% (28/39), 82.1%
(32/39), and 79.5% (31/39) were respectively observed among an
early infection category of “Acute” samples. Of the vendor-tested
Lyme positive samples, positive agreements of 68.6% (72/105),
52.4% (55/105), 61.9% (65/105), 64.8% (68/105), and 70.5%
(74/105) were observed for the BioPlex 2200 Lyme Total, DiaSorin
LIAISON, Immunetics C6, Zeus PepC10/VlsE1, and BioMerieux
VIDAS Lyme II, respectively. Additionally, positive agreements of
69.4% (50/72), 55.6% (40/72), 58.3% (42/72), 62.5% (45/72),
and 73.6% (53/72) were respectively observed for the acute group
of vendor samples.
In parallel, non-endemic prevalence values of 3%, 4%, 4%, 7%, 10%
and endemic values of 5%, 3%, 8%, 6%, and 5% were observed for
the BioPlex 2200 Lyme Total, DiaSorin LIAISON, Immunetics C6,
Zeus PepC10/VlsE1, and BioMerieux VIDAS Lyme II.
Conclusion: Overall, the BioPlex 2200 Lyme Total assay exhibited
an unmatched combination of both relatively high sensitivity and
specificity that was not observed among commercially available
Lyme kits. Importantly, the improved sensitivity of the BioPlex
2200 Lyme assay was primarily from early Lyme disease samples
observed in both the CDC Premarketing panel as well as vendor-
supplied samples.
Presenter: Johnnie Hahm, Bio-Rad Laboratories, Hercules, CA,
[email protected]
PublicHealthLabs
@APHL
APHL.org
Methods: Galileo AMR™ HT was used to annotate all of GenBank.
The nucleotide database from NCBI was downloaded (Jan 28,
2018). BLAST was performed for all sequences in the Galileo AMR™
HT database except for very common elements such as IS1. The flat
file for accessions that had one or more matches was subsequently
downloaded[1] and species names extracted. All synthetic,
eukaryotic and non-CMV viruses were removed. The sequences of
the remaining ~22,000 sequences were converted to FASTA format
and annotated using Galileo™ AMR. Annotations in the flat file were
ignored.
Background: Lyme disease is a growing public health concern as
the number of US counties reporting its incidence has increased
substantially over the past decade. Early detection of Lyme disease
is of high clinical value for treatment efficacy and the prevention of
more serious, long-term health consequences. Bio-Rad Laboratories
has developed a multiplexed Lyme assay that detects both IgM and
IgG antibodies using a unique combination of synthetic peptides
and recombinant protein antigens for an improved early detection
without sacrificing high specificity.