FROM THE BENCH
As the clinical diagnostics industry moves away from isolates toward
genomics, we need to change our surveillance so that we can be prepared
for an era when there are no more isolates. One way to do that is to develop
genomics-based testing on the public health side.”
• Engage your quality assurance officer early
in the process to help identify relevant
metrics.
Kristy Kubota, MPH
First Steps
Matluk was tasked with finding a process
that would provide the most functionality
given the lab’s size, expertise and
resources. Procuring the next-generation
sequencer was one of the first hurdles.
The group was able to buy the equipment
using Epidemiology and Lab Capacity
funds, but the state purchasing process
took nearly a year.
Training on the new wet lab protocols
was relatively straightforward for the
laboratory staff, he said, but the in silico
methods posed a somewhat larger
challenge.
“We don’t have a dedicated
bioinformatician or dedicated IT staff,”
Matluk said. “So I really need a nice, easy,
graphical user interface that looks like
any other web application or computer
program that you can easily write CLIA-
validated protocols for. ‘Click the upload
button, click the workflow button’ is much
easier to write into a CLIA protocol than
command line code.”
After considerable research, he chose
Qiagen’s CLCbio software as the lab’s
primary bioinformatics platform, which
has worked well for in-house analyses. It
is fast, user-friendly, and Matluk likes that
he can develop and preset metrics, then
lock the workflows for quality assurance
(QA) purposes.
Choosing and setting those QA metrics
was the most critical and time-consuming
step. “We went from knowing nothing to
trying to learn a lot, so I spent the better
part of about six months just figuring
out what we wanted to monitor, how to
monitor and what actions to take when
certain metrics were failed,” Matluk said.
PublicHealthLabs
Nicholas Matluk shares
some advice for other
groups starting out:
@APHL
Once the metrics were set, CLIA validation
took about four months of work time,
which the staff spread over the course of
a year to fit alongside full-time PFGE work.
To streamline the process, they validated
the wet bench and in silico components
separately and focused on validating the
protocols against their existing databases
of bacterial sequences rather than against
individual genes.
Overall the process has gone smoothly,
Matluk said, with the backing of the lab
director and state infectious disease
epidemiologists. He also needed buy-in
from the state Office of Information
Technology for assistance procuring
the needed computer equipment
and developing security protocols for
connecting to networks to upload and
download information.
Three years since starting the transition,
the Maine lab now has four employees
PulseNet-certified on six organisms for
sequencing, QA/QC metrics and reporting
out results.
Broader Impacts
With several WGS protocols now
validated, the Maine group has found
that they can explore more questions
than they did with PFGE. In addition
to species identification, they perform
in silico serology for several strains and
multilocus sequencing typing, as well as
analysis for antibiotic resistance genes
and virulence factors. They have also
validated their WGS protocols and some in
silico pathways for healthcare-associated
infections, antibiotic resistance genes and
phylogenetic tree analysis using single
nucleotide polymorphisms.
“Once you have the sequence data, you
can pretty much use it however you
want as long as you have that workflow
APHL.org
• Have at least two or three points of failsafe
data backup.
• Engage your epidemiologists to determine
which testing methods can be replaced
by whole-genome sequencing and which
should not to ensure actionable infection
control.
• Engage your LIMS administrators to help
develop lab reports.
or pathway validated,” Matluk said. For
example, the phylogenetic tree analysis
has helped with cluster outbreak analysis,
while an incidental findings protocol,
modeled after one developed by the
American College of Medical Genetics,
reports out findings of interest—such
as specific genes or plasmids, or certain
resistance patterns—that were not part of
the immediate analysis but may inform
larger investigations.
The additional information Maine
and other labs are generating with
WGS is especially valuable for national
surveillance efforts through PulseNet,
Kubota said.
Sevinsky sees WGS as a step within
an even bigger transition to using
metagenomics-based tests in public
health. WGS databases could allow direct
querying of a specimen from a patient to
get faster and more complete answers: Is
the organism related to a cluster? Does it
have specific antimicrobial resistance or
virulence factors?
“That’s still a couple years in the
future,” Sevinsky said. “But the work
that we’re doing right now with whole-
genome sequencing is going to lay the
foundation for the next generation of
metagenomics.” n
Fall 2019 LAB MATTERS
25