Co-activation probability between neurons in the largest brain connectome of the fruit fly

This is a data set containing the co-activation probability between neurons in the largest brain connectome of the fruit fly released by the FlyEM project. The co-activation probability is measured based on neural dynamics computation, where a standard leaky integrate-and-fire (LIF) model is applied on the connectome to generate neural dynamics. Please read the paper "Yang Tian, Pei Sun; <strong>Percolation may explain efficiency, robustness, and economy of the brain</strong>. <em><em>Network Neuroscience</em></em> 2022; 6 (3): 765–790. doi: https://doi.org/10.1162/netn_a_00246" for more details. <strong>This is the newest version of the data set.</strong> The following is a list of variable information: (1) SomaLocation is a 23008*3 matrix that contains the three-dimensional coordinates of neurons; (2) LambdaVector is the vector of a vector of synaptic excitation-inhibition (E/I) balance (see "Percolation may explain efficiency, robustness, and economy of the brain" for detailed explanations). (3) DirectedCoactivationPattern is a cell of co-activation probability matrices generated under each E/I balance condition, which is used in "Percolation may explain efficiency, robustness, and economy of the brain" for computational experiments. The (i,j)-th element in the matrix is the probability for neuron i to activate neuron j under the corresponding E/I balance condition. Note that the (i,j)-th element can be differnt from the (j,i)-th element. (4) SymmetricCoactivationPattern is a cell of symmetric co-activation probability matrices generated under each E/I balance condition. This is a new data that has not been used in "Percolation may explain efficiency, robustness, and economy of the brain" yet. The (i,j)-th element in the matrix is the probability for neurons i and j to be co-activated under the corresponding E/I balance condition. If we define D as the directed co-activation probability matrix and denote S as the symmetric co-activation probability matrix, then we have S(i,j)=S(j,i)=0.5*(D(i,j)+D(j,i)). The earlist version of this data can be seen in https://zenodo.org/record/5497516, which may lack detailed explanations. The second version of this data can be seen in https://zenodo.org/record/7869532, where a small mistake is found while calculating the SymmetricCoactivationPattern. This mistake is resolved in the newest version.