Biomedical Engineering Reference
In-Depth Information
expected hit and is also RNAi mediated. This further helps in
assessing transfection effi ciency. Negative controls should be
non-targeting, have little cytotoxicity and result in no change
in phenotype from untransfected cells. There should be a suf-
fi cient number of control wells per plate for statistical analysis.
6. Another critically important experimental factor is the cell count-
ing, as it impacts both transfection effi ciency and cell behavior.
Be scrupulously careful when counting cells. Perform repli-
cates, and use the same counting method once established.
Also try to count cells within a suitably optimal range of cell
density, because different cell counters may have different
optimal working ranges.
7. To minimize data variability, pick a transfection reagent whose
performance is stable over time and across batches, if possible.
It is best practice to test and then purchase a large batch of the
same lot number of reagent.
8. A commonly used metric for assessing the quality of a high-
throughput assay is the
Z
-factor, given by 1 − 3(
˃
p
+
˃
n
)/|
ʼ
p
+
ʼ
n
|,
where
˃
and
ʼ
represent respectively the standard deviations
and means of the positive (
p
) and negative (
n
) controls. It gives
a measure of the variability and dynamic range between the
positive and negative controls. A
Z
′
-factor showing very good
separation of controls should be between 0.5 and 1; however,
for siRNA screens, a
Z
′
-factor above 0 may be acceptable [
33
].
9. Common plate position artifacts that may be encountered
include edge effects due to evaporation from assay plates and
row and column artifacts due to liquid handling steps and/or
chemical stability. These can be minimized by dedicating incu-
bators for the screen (avoiding constant opening and closing of
doors to maintain a consistent environment) small volumes of
liquid, and careful handling of reagents (e.g., not leaving light-
sensitive reagents exposed to light while in liquid dispenser
tubing).
10. Cellular heterogeneity may require removal of artifacts wher-
ever possible, such as out-of-focus fi elds of view, apoptotic and
mitotic cells, or cells on the border of the image. Alternatively,
averaging of a large number of cells may dilute and negate the
effect of the artifacts.
11. Controls should be systematically distributed on the plate, to
minimize skewing of readout due to plate positional artifacts.
12. Excess transfection reagent mix/cell suspension should be
prepared, to account for both the liquid handler dead volume
and general liquid loss during transfer. The amount of excess
is best determined empirically during the optimization stage.
′
