Wrong hoc templates in models in Hippocampus models container

[sasaray]

The PLOS paper models in the container: https://object.cscs.ch/v1/AUTH_c0a333ecf7c045809321ce9d9ecdfdea/hippocampus_optimization/ contain hoc templates that still require the axon to have at least 2 sections. The problematic lines are in "proc replace_axon()"

     access axon[0]
     i1 = v(0.0001) // used when serializing sections prior to sim start

     access axon[1]
     i2 = v(0.0001) // used when serializing sections prior to sim start

These lines are commented out in the new models, and I think were commented out in the GitHub versions of the PLOS paper models. As in some morphologies the axon has only 1 section, we can not run, and validate those models. This is blocking us with @appukuttan-shailesh in creating a live paper on HippoUnit.

@mmigliore, @rmiglior, @clupascu: Could you please replace the hoc templates with the correct ones in the container?

[clupascu]

@sasaray the PLOS paper models are in https://object.cscs.ch/v1/AUTH_c0a333ecf7c045809321ce9d9ecdfdea/Migliore_2018_CA1/models/ not in https://object.cscs.ch/v1/AUTH_c0a333ecf7c045809321ce9d9ecdfdea/hippocampus_optimization/. Can you please check that the hoc templates in the PLOS paper models container above suits your requirements?

[appukuttan-shailesh]

@clupascu .... there was a confusion. @sasaray is infact referring to the container: https://object.cscs.ch/v1/AUTH_c0a333ecf7c045809321ce9d9ecdfdea/Migliore_2018_CA1/models/ (and not /hippocampus_optimization/)

The URLs had been picked up from the model catalog, and there we redirect to "Migliore_2018_CA1" container. e.g. https://collab.humanbrainproject.eu/#/collab/8123/nav/395557?state=model.b1a62b3e-55ea-4b23-939f-94862ad56516 gives the following URL: https://object.cscs.ch/v1/AUTH_c0a333ecf7c045809321ce9d9ecdfdea/Migliore_2018_CA1/models/CA1_pyr_cACpyr_oh140807_A0_idA_20170915112548/CA1_pyr_cACpyr_oh140807_A0_idA_20170915112548.zip

[appukuttan-shailesh]

p.s. some of the models known to have this issue are: CA1_pyr_cACpyr_oh140807_A0_idA_20170915112548 CA1_pyr_cACpyr_mpg141215_A_idA_20170915152442 CA1_pyr_cACpyr_oh140807_A0_idF_20171003112912 CA1_pyr_cACpyr_oh140807_A0_idG_20170915113354 CA1_pyr_cACpyr_oh140807_A0_idC_20170912112759 (the others seemed to be ok, but you could double-check)

[clupascu]

ok. we will fix it.

[clupascu]

I fixed CA1_pyr_cACpyr_mpg141215_A_idA_20170915152442. Can you please test that this model works for you? I actually noted that all models have this issue, not only the ones listed by @appukuttan-shailesh above.

[sasaray]

Yes, all of them has this in the hoc template, but in case of those models, that have at least 2 axonal sections, this doesn't cause any error.

[clupascu]

I fixed all the models in the PLOS paper container. Please check.

[appukuttan-shailesh]

Thanks @clupascu for the quick work on this. I tested a few of them and they worked. Will review all and update this ticket.

[appukuttan-shailesh]

I am having a problem with "CA1_int_cNAC_970509HP2_20180117154753" model.

Error:

NEURON: /media/shailesh/DATA3/Downloads/CA1_int_cNAC_970509HP2_20180117154753/morphology/970627BHP1.asc :file is not open

It's hoc files require to load the above morphology file "970627BHP1.asc":

  if(numarg() >= 2) {
    load_morphology($s1, $s2)
  } else {
    load_morphology($s1, "970627BHP1.asc")
  }

But the morphology directory of this model only contains "970509HP2.asc".

@clupascu : Could you fix this if it is indeed an error?

[clupascu]

@appukuttan-shailesh I fixed this. It was indeed an error.

[appukuttan-shailesh]

Thanks @clupascu ! All the problems with the live paper models have been resolved. Closing this issue now.