2. We found that on the timescale of the entire assay, it was
unnecessary to wait for a hypomorphic phenotype to develop
after removal of ATC: results were similar whether waiting
0, 12, or 24 h. The lowest inoculum we tested, equivalent to
an optical density at 600 nm (OD 600 ) of 0.0025, gave the
greatest dynamic range, presumably because the difference
between inoculum and carrying capacity was greater. Finally,
we tested incubation times of 7, 10, and 14 days, with the
longest of these producing the greatest dynamic range.
3. We found that adding a solution of 10% (v/v) DMSO in water
at 1:1 volume ratio with the bacterial culture followed by
extensive pipette mixing (to break up the bolus of
M. tuberculosis that forms after 14 days of standing growth)
and heat released enough genetic material for reproducible
library construction. Carryover of DMSO to the downstream
PCR reaction likely assists in amplification efficiency. At this
step, we also add to the DMSO stock solution, a low concentration “spike-in” plasmid containing the same barcode cassette (but a distinct barcode) as the strains. Therefore, the
spike-in plasmid can act as a tracer of variability of downstream
molecular biology independent of the biological variability of
the strains. While the utility of spike-in plasmids for quantitation is arguably limited [11], their barcode counts (or lack
thereof) can be used to troubleshoot problems in downstream
processing. They can also be used in sanity checks when calculating biologically relevant quantities, such as strain fitness:
because the spike-ins do not grow, they should have a fitness
close to zero.
4. In our experience, black polystyrene will warp during the 80
C
decontamination step.
5. Array test compound stock solutions at an appropriate volume
in the assay plates beforehand, either by pipetting or by using
an acoustic dispenser. The outermost wells of each plate can
give unreliable results, so these should not be used.
6. We double-bag assay plates in autoclave bags, sealed tightly
with rubber bands.
7. Both spike-ins are plasmids containing the same PCR primer
binding sites to enable incorporation into the final sequencing
library, but contain a distinct strain barcode so they can be
distinguished. They are added at to a final concentration of
2.6 fM, approximately the same concentration as that of each
strain’s chromosome at the end of the assay. This ensures that
the spike-in controls are detectable, but do not unnecessarily
consume sequencing reads.
8. We use 96 forward primers incorporating 5
0 -overhangs containing the same Illumina P7 sequence for flow cell hybridization, but which are unique in a barcode indicating one of the
340
Eachan O. Johnson and Deborah T. Hung
Précédent

- 347/734

Suivant