These scripts were used for supplementary Figure S5.
Because there were some inaccuracies in the figure caption,
the rationale is explained in greater detail here.

Panel A tries to repeat the findings of Smulders and Zwart
that co-agonism at the ACTIVATION gate can cause an amplification
of the response to weak ACh pulses (upper panel).
The bath compound (a so-called competitive antagonist) has itself 
zero efficacy, but when it co-binds with ACh it opens the receptor.
Further it is assumed here that the applied ACh does not desensitize the receptor
(this assumption is probably wrong, it would be valid for physiologically
released ACh, but not for ACh applied in 5 sec pulses).
When the ACh pulses are taken stronger (60 microM), the facilitating
effect of the antagonist completely disappears, and becomes even depressing,
as can be expected for a competitive antagonist (bottom panel).

Panel B gives an explanation for the findings by Prickaerts et al.
that very low doses of a partial a7 agonist amplify the responses
to ACh pulses. Here the compound (EVP-6124) is a partial agonist
with an efficacy of 0.3. At a low dose (upper panel) the compound
amplifies the ACh response through co-agonism at the DESENSITIZATION
gate, hence co-binding of EVP-6124 and ACh prevents desensitization 
of the a7 receptor. (Obviously in this simulation the ACh pulses
do desensitize the receptor, as can be seen from the initial transient.)
When a high dose of EVP-6124 is applied in the bath (lower panel), 
however, binding of two EVP molecules occurs much more frequently 
than co-binding of EVP and ACh, so that the a7 receptor becomes 
strongly desensitized and the ACh response drops as fewer receptors are available.