Modeling
Genetics Screens by Retroviral cDNA Transfer: Example Screen #1
This
page demonstrates a genetic screen in which a near ideal phenotypic
selection occurs.
-
The
graph and explanations are excerpted from a down-loadable demonstration
Microsoft Excel 5.0 file written by GPN for modeling genetic screens.
-
The
equations in the Excel file are relatively primitive, do not take
into account statistical representations of population dynamics,
and are meant as a first time trial to help you think more intuitively
about genetic complementation cloning by retroviral gene transfer.
-
Yes,
there are more complicated versions of the file that one could write,
and you are welcome to modify the equations in the spreadsheet after
download.
-
If
you have any interesting additions, please let me know and I will
put them up on the site here and credit your brilliance.
-
For
details on the use of these files please see the Excel File on download
and documentation contained therein.

In this example
the real event sought is set at 1 per million;
The heritable
background is set at 1 x 10E-5;
The sporadic
background is set at 1 x 10E-5 as well.
The efficiency
of the rescue is set to 50%.
As expected,
within two selection/generations you have selected for a population that
still reflects the ratio of real to false events, but in which the real
event is now nearly 5% of the total population and worth attempting
to clone directly.
Note that
if you rescue the insert and re-infect that you will get nearly
50% of the population (light blue squares) positive at the second round
of selection.
-
The graph
as presented here is relatively primitive. For example, the "rescue"
graph (light blue squares) only considers the events from the previous
generation and are therefore not connected by any line.
-
The graphs
also assume that the rescue frequency is the same as the initial infection
frequency upon transfer of the library to the cells at Gen_0.
-
The other
major assumption in these graphs and equations is that only a single
virus occupies each cell. More complicated equations requiring Poisson
distributions can be applied if you want to rewrite the spreadsheet.

The
numbers in the chart below (D20 through I20) are used to generate the
graph above. There is usually no need to change either the numbers within
the chart directly or to change the equations for each cell position.
The numbers are derived originally from what you type in positions B1,
B3, B5, and B7.
|