post database analysis

[AmosG8]

1. adjust new fields within the structure to match the new EP naming.
2. It used to load the dataset (EP_FILES_COMPILED) which was constant and now it has a different name and it's dynamic.
3. Add juxta condition to AG_Summarize day's output.
4. Analyze the output of AG_summarizeDay - line 2 of the cell array is the combined data of all experiments. We want to quantify its mean and perhaps kurtosis.
5. For every stimulus in the data find the average DF/F in a time window before and after that stimulus. We can also do it with spontaneous data by randomizing kernels during the experiment.
6. Correlation analysis with AG_analyzeCorrCompiled (use my flowchart if necessary). -expressed in bins according distance for each fov and then averaged across a specific day.
7. The graphs that we want to have at the end are: -a comparison of hypo to hyper across days for the same condition (for example standing during stimulation).
   Parameters to analyze: Magnitude of response: -we will want to average the dF for each cell for the standing stim frames then to average that across all cells in that imaging day, This is a data point on the graph.
   • Fidelity -we'll need to decide what is a response and then find the percent of times the cell responded (lets say 7 out of 10 is a fidelity of 70%). -Percent of responsive neurons -we'll need to define each neuron as responsive or not. For this we will test if its evoked activity significantly higher than its spont activity. so each cell receives either 0 or 1. Then we analyze what is the percent of responsive neurons per field of view. Then, we can re-analyze the magnitude of response and fidelity taking in account only responsive neurons.

[HagaiHargil]

More things to consider:

1. Spike amplitude.
2. Spike frequency.
3. AOC. Consider more parameters from analysis done in the field, like Konnerth, Jasor Kerr.