Brendan Frick

Agent Behavior:

	- I've refined the movement and I think what I have is what I'll use in my final project. I've reduced step
          size so agents don't jump over membrane. I also perform a check sweep every iteration to remove ions that somehow jumped 
          without passing through a channel. By going through a channel, a location variable switches in the ion. 

        - I'm starting to play with channel selectivity, all the channels are selective to one ion (which is more or less true in the cell).
	  I'm working on one way current selectivity.

Rationale for agent rules:

 	- Movement looks about right. Since movements in jumps, sometimes I will get errors, this reflects an error in the 
	  regression from real life to virtual world (where barriers don't exist). My check sweep to remove ions reflects 
	  the impenatrability of real life membranes.

Model Output
	
	- I now have a voltage plot. The voltage plot wavers around a constant number which is indicative of 
	  equilibrium random movement (which is what I want right now). Once I start adding voltage gating properties, I should 
	  see some recognizable trends.

Questions
		
	- How should I implement the influence of current flow. In the cell, when the membrane voltage is polarized (either too high
	  or too low), there's a massive shift in the concentration gradient and ionic current increases either in or out. I have 3 options 
  	  that I could think of - 1) force 2) probability 3) range

	1) Force would be calculating the concentration gradient for each transmembrane path. This is the real life predicament but I think
		it would be too messy in Netlogo. I am leaning towards the second or third option
	2) Every time an ion gets a chance to pass through a membrane, it has a probability, based on the concentration gradient, to pass 
	   through. This idea might work but I'm afraid that if I restrict transmembrane travel we won't have a fluid voltage (it already 
	   is difficult enough to attract ions to channels because I don't have a force component.

	3) Range. The bigger the gradient the more range this channel has. The minimum range is neighbors, second tier range (medium gradient)
		is neighbors of neighbors, highest gradient would be neighbors cubed. I'm going to start with this one, and then try probability if it doesn't work.

Model Analysis

	Models coming along. At this point I need to be able to implement channel selectivity and voltage gating and I should have a 
	model of a fundamental neuron.
	
	- None so far.
