Spike source and target sections

A Python package that facilitates development and use of models of biological neural networks

Moderator: tom_morse

bremen
Posts: 26
Joined: Mon Apr 24, 2017 8:15 am
Location: Italy

Re: Spike source and target sections

Post by bremen » Sun Mar 25, 2018 7:34 am

Hi Salvador.
I'm checking that everything, so far, is correct but, looking at the complete output, i saw something strange:

Created connection preGid=0, postGid=1, sec=dend_0, loc=0.7421, synMech=CF_AMPA, weight=1, delay=1.00, threshold=10
Created connection preGid=0, postGid=1, sec=dend_0, loc=0.2263, synMech=CF_AMPA, weight=1, delay=1.00, threshold=10
Created connection preGid=0, postGid=1, sec=dend_1, loc=0.8218, synMech=CF_AMPA, weight=1, delay=1.00, threshold=10
Created connection preGid=0, postGid=1, sec=dend_1, loc=0.5042, synMech=CF_AMPA, weight=1, delay=1.00, threshold=10
Created connection preGid=0, postGid=1, sec=dend_1, loc=0.1866, synMech=CF_AMPA, weight=1, delay=1.00, threshold=10
Created connection preGid=0, postGid=1, sec=dend_2, loc=0.5578, synMech=CF_AMPA, weight=1, delay=1.00, threshold=10
Created connection preGid=0, postGid=1, sec=dend_3, loc=0.08891, synMech=CF_AMPA, weight=1, delay=1.00, threshold=10
Created connection preGid=0, postGid=1, sec=dend_4, loc=0.146, synMech=CF_AMPA, weight=1, delay=1.00, threshold=10
Created connection preGid=0, postGid=1, sec=dend_6, loc=0.7226, synMech=CF_AMPA, weight=1, delay=1.00, threshold=10
Created connection preGid=0, postGid=1, sec=dend_7, loc=0.4523, synMech=CF_AMPA, weight=1, delay=1.00, threshold=10
Created connection preGid=0, postGid=1, sec=dend_8, loc=0.6848, synMech=CF_AMPA, weight=1, delay=1.00, threshold=10
Created connection preGid=0, postGid=1, sec=dend_8, loc=0.04196, synMech=CF_AMPA, weight=1, delay=1.00, threshold=10
Created connection preGid=0, postGid=1, sec=dend_10, loc=0.4802, synMech=CF_AMPA, weight=1, delay=1.00, threshold=10

Counting all the lines are exactly 105 connections, as defined in the seclist and in synsPerConn, but there are repeated sec name and sec missing.
The sec locations, instead, are all different, so i can think that is only a name problem, since the physiological response is correct.

The section list i'm passing does never repeat but it has some gaps after the first 70 values.
cf_dend = ['dend_0', 'dend_1', 'dend_2', 'dend_3', 'dend_4', 'dend_5', 'dend_6', 'dend_7', 'dend_8', 'dend_9', 'dend_10', etc...]

Best
Stefano

bremen
Posts: 26
Joined: Mon Apr 24, 2017 8:15 am
Location: Italy

Re: Spike source and target sections

Post by bremen » Sat Apr 28, 2018 3:51 pm

Hi Salva.

I have a doubt about this for cycle.

Code: Select all

pop_C = 10
for p in range(0,len(pop_C)):
    netParams.connParams['Pop_B->Pop_C'+str(p)] = {'preConds': {'pop': 'Pop_B'}, 'postConds': {'pop': 'Pop_C'},
            'weight': 1, 		
            'sec': location[p],
            'delay': 0.1,					
            'loc': 0.5,					
            'synsPerConn' : 30,
            'connList': [Pop_B_2_Pop_C_section_cell[p]],
            'synMech': 'Syn_B_C'} 
For example. If p =0, the connection will be made from cell_B [12] to cell_C [0] (connList), with 30 synapses for connection, concentrated on a single section of cell_C [0]. The next one, p = 1 will be the same but with a different connList, for example [3,0] and so on, everytime with a different connList.

My doubt is, am I actually placing 30 synapses for each connection, for a total of 300 synapses? Or i'm doing 10 connections which overwrite the one before and i have, at the end, only 30 synapses?

Best
Stefano

bremen
Posts: 26
Joined: Mon Apr 24, 2017 8:15 am
Location: Italy

Re: Spike source and target sections

Post by bremen » Sat May 12, 2018 12:07 pm

Hi.

I have changed the previous code with something simpler but there is still something odd.

Let assume to connect 10 B_cell to 1 C_cell on the same section ['soma'] with 10 synPerConn.

Code: Select all

    netParams.connParams['B_pop_2_C_pop] = {'preConds': {'pop': 'B_pop'}, 'postConds': {'pop': 'C_pop'},	
            'sec': ['soma'],						
            'synsPerConn' : 10,
            'connList': [[0,0], [0,1],etc,] 	
If 'sec' is written as a single section ['soma'], the synaptic impact is stronger then if 'sec' is written as a list composed by 10 repetition of the same section ['soma', 'soma', etc]. Using verbatim i have checked the connections and, in both cases, each connection is generated on the soma.

Cheers
Stefano

Post Reply