Join GitHub today
GitHub is home to over 28 million developers working together to host and review code, manage projects, and build software together.Sign up
This code is the basis of the probabilistic model and the biophysical model in the publication by Field et al. 2018: Heterosynaptic Plasticity Determines the Set-Point for Cortical Excitatory-Inhibitory Balance, doi: https://doi.org/10.1101/282012
Fetching latest commit…
Cannot retrieve the latest commit at this time.
|Failed to load latest commit information.|
This file lists all the parameter-values used for the simulations in the publication: Heterosynaptic Plasticity Determines the Set-Point for Cortical Excitatory-Inhibitory Balance (2018) Rachel Field, James D'amour, Robin Tremblay, Christoph Miehl, Bernardo Rudy, Julijana Gjorgjieva, Robert Froemke bioRxiv, doi: https://doi.org/10.1101/282012 The code was written by Christoph Miehl (firstname.lastname@example.org). The concept was developed by Christoph Miehl and Julijana Gjorgjieva (email@example.com). Date: July 2018 For the probabilistic model in Figure 3A-C: Figure 3A Parameters are defined as: channels=12; eLT=1.65; iLT=1.65; ehet=0.62; ihet=0.62; Figure 3B (upper panel) Parameters are defined as: channels=12; eLT=1.65; iLT=1.65; ehet=0.62; ihet=0.62; Figure 3B (lower panel) Parameters are defined as: channels=12; eLT=1.65; iLT=1.65; ehet=0.2; ihet=0.2; Figure 3C Here the heterosynaptic decrease is changed systematically. The figure is generated for ehet=[0.98:0.02:0.02] and ihet=[0.98:-0.02:0.02] where the two values are the same in each simulation ehet=ihet. Note that the ratio in Fig. 3C is calculated by dividing the heterosynaptic weight decrease by the homosynaptic weight increase. So for example, when ehet=ihet=0.98, then the ratio of heterosynaptic (100%-98%=2%) to homosynaptic strength is 2%/65%=0.0308. For different numbers of channels (e.g. 6), the parameter "channels" has to be changed accordingly. For the biophysical model and Supplementary Figure 4: Figure 3D-E Parameters are as defined in the file "Biophysical_Model.m". An instance of this code runs 1-2 hours on a MacBookPro (2014 model). Figure 3F To titrate the learning rate ratio of heterosynaptic vs homosynaptic plasticity, the heterosynaptic learning rates are changed systematically. The figure is generated for eta_het_E=[0.1,2.5]*10^(-5) and eta_het_I=[0.1,2.5]*10^(-4). Note that both values are changed in the same way, if eta_het_E=0.1*10^(-5) then eta_het_E=0.1*10^(-4). The ratio of eta_het_E/eta_het_I=10^(-1) is always the same. To generate the full Fig. 3F requires writing an additional loop, which will require this code to be run a number of times (each 1-2 hours duration). All other parameters are as defined in file "Biophysical_Model.m". For different numbers of channels, the parameter "channels" has to be changed accordingly. Figure S4B Parameters are as defined in the file "Biophysical_Model.m".