X-Message-Number: 26933 From: Date: Mon, 5 Sep 2005 14:33:59 EDT Subject: Uploading technology (1.i.2) The post synatic part. Uploading technology (1.i.2) The post synatic part. There are symetrical, inhibitory synapses on dendrites, soma or the first part of the axon, the hillock. The other class, the asymetric synapses include all the excitatory ones, mostly working with glutamate or aspartate. These are often on dendritic spines. There is no spine without an excitatory synapse, even if 10 15 percent include too a symetric, inhibitory component. The difficulty, for a simulation come from the channel diversity. The synaptic functions to be included at the postsynaptic level are: -The channel famillies. -The spine. -The receptor management. Information input is not limited to the presynaptics source: Neurotransmitters come from both, the presynaptic element and a more diffusive component from other synapses or gap junctions. This is particularly true for second transmitters. There are three potential signals from the dendrite: The back propagating axion potential, the nearby synapses potential and the ionic gap junctions on dendrites. All of that can be summed up as a dendrite potential. Because of the small capacity of the synapse membrane, there is no need for a separate work on slow and fast varying potentials. The dendrite input potential (DIP) is used such that on flat synapses and must be processed for spines. Depending on the spine shape the potential will not be the same at the head. So the DIP is processed according the spine index (SInd). Fast channels, tension gated or main neurotransmitter gated ones are not modulated by second or third transmitters on asymetric synapses, there is so a symetry synapse index (SymSInd). For example SymSInd = 0 for excitatory synapses, so all neuromodulator effect is cancelled here. If SymSInd = 1, then this effect must be taken into account. A post synaptic head contains one thousand or more channels, this is sufficient to smooth out any individual statistics and there is no need for stochastic mathematical methods. That let neverthless a lot of parameters from the channel diversity. G-protein and second messenger gated channels are computed at the general computer level and enter here as a single tension input, (CompTIn) included in the Computer Input (CompIn). CompIn may not be actualised at each computing cycle, so this parameter could be used for a number of synapses, at each cycle, CompIn would refresh the data in a given synapse, comming back to it after 128 cycles or more. CompIn would include informations such as produce or destroy x percent of a given receptor or act on spine shape. On the output side, there are 3 signals: - The potentials sent to the dendrite, one for each channel species (PotOut). - The back potential sent to the presynaptic part, using the gap junctions and capacitive effect (BackPot). - The NO/CO gas sent to the presynaptic part and nearby synapses. This is modeled by a signal sent to the computer level, the computer out (CompOut) answer to CompIn. So we have on the input side: The channels parameters, DIP, SInd, SymSInd, CompIn including CompTIn. On the output there are: PotOut (on the order of 15 currents), BackPot, a single value and CompOut. Yvan Bozzonetti. Content-Type: text/html; charset="US-ASCII" [ AUTOMATICALLY SKIPPING HTML ENCODING! ] Rate This Message: http://www.cryonet.org/cgi-bin/rate.cgi?msg=26933