Hope for the best...

because the Interaction Trap yielded on average four times
more successes than the failures. But you have to know when prepare to
the worst!
We assume that we have a representative set of data. Which may be not true
IF the most dissappointed researchers did not report us their failures...
Major traps for a hunter
In our opinion, the two biggest dangers preventing success of the two-hybrid
hunts are transactivation by the bait protein itself and failure
to express the bait properly. For these obvious reasons a few researchers
abandoned hunts without even starting a screen. Among about analyzed
115 library screenings, 16 were reported to be total failures (i.e. either no positives, either only false positives). These cases can be tentatively explained as follows:
- 4 colleagues experienced difficulties with protein expression (i.e. no protein was detected on Wester blots);
- 3 failures could be due to weak transactivation by the bait protein;
- in 2 cases, a number of primary transformants too low to represent genome complexity is suspected;
- in one case the transformants were plated
directly on selective medium without prior plating on UHW medium;
- finally, 2 screening were performed with all possible precations, and
no obvious reason can be evoked to explain the failure in these cases.
However, it can be either the absence of the alleged interactor in the
library, or even more profound biological reasons.
- 4 colleagues were sufficienly disappointed that they did not answer the questions which could give us clue on the possible reasons of these particular failures
Your chances of success
Analysing the bulk of the reported screenings, and assuming that our database is representative, we can conclude that:
- if your protein is properly expressed and is not activating either
of the reporters, odds are 6 to 1 that you will pull out something which makes biological sense, if you screen an adequate number of clones.
Note: for human, genome complexity will be represented by approx. 1-2 million primary transformants;
for Drosophila, by ???.
- if your protein is weakly transactivating, your chances drop, but not
really dramatically. However, if your protein is strongly activating leu
reporter, we would recommend refraining from plating directly on X-Gal plates.
Three colleagues attempted, and none succeeded.
- if you cannot detect your bait protein in yeast, your chanses drop
substantially. Consider the situation once again before venturing on a screen.
As in case with activating baits, you can spend a lot of time for
nothing.
Condi- tions
|
No activation, protein expressed
|
Weak activation or leu activation
|
Problems with the protein expression
|
red: +
success
blue: -
failure
green:
? ongoing |
 |
 |
 |
Total:
|
57
|
18
|
10
|
The degree of your success can vary!
The complexity of many genomes, and the complexity of the web of
protein interactions is still beyond of the abilities of any human-made
systems. So, you will find not necessarily what you want to find;
rather, it is better not to to be ruled by preconceptions and to be aware of the limitation
of the system. This can be illustrated by following data:
of 64 successful hunts, the partners were known for 30
baits before screenings.
However, only 11 of these baits found previously known partners, whereas 19 baits only identified novel, albeit biologically relevant partners.
This analysis emphasizes that the two-hybrid systems are the
tools to uncover unanticipated interactions!