Central repository of .prb files

122 views
Skip to first unread message

Max Hunter

unread,
Dec 2, 2015, 1:48:27 PM12/2/15
to phy-users, klusta...@googlegroups.com
Dear all,

I've created a central open-source repository at https://github.com/kwikteam/probes/ to contain probe files from different manufacturers. Please contribute your own!

In future phy will have an option to download this repository automatically and keep a local copy (so new probes/manufacturers will be added as they come out, and you can just specify the probe name in the .prm file), but for the moment, I would really appreciate any contributions you all have - at the moment Justin Kiggins has very kindly contributed an automatically generated set of all the NeuroNexus linear probes (1028 of them!) but if anyone has any .prb files from other manufacturers, we'd really appreciate your contribution!

If you're unsure of how to make a GitHub pull request, please drop me a message and I can help or include them directly myself.

A note from Justin about the NN probes:
"these are the probes in the nn database labeled "linear". I think some are actually custom, however, and are not accurate. further, I couldn't parse the logic for tetrodes or poly-trodes."

A discussion on automatically generating probe files is here: https://github.com/kwikteam/phy/pull/521.

In the future, we will probably have an option to generate the connectivity graphs automatically (see http://www.justinkiggins.com/blog/generating-channel-graphs-for-klustakwik-automatically/) but in any case, the geometries will still be manually needed - if possible please specify these in microns so generating the connectivity graphs will be easier (and so other algorithms which need the physical probe dimensions can take advantage of the same probe files)

Best,
Max

Brendon Watson

unread,
Dec 2, 2015, 2:03:53 PM12/2/15
to Max Hunter, phy-users, klusta...@googlegroups.com
Thank you

--
To subscribe, send email to klustaviewa...@googlegroups.com
---
You received this message because you are subscribed to the Google Groups "KlustaViewas" group.
To unsubscribe from this group and stop receiving emails from it, send an email to klustaviewas...@googlegroups.com.
To post to this group, send email to klusta...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

Vahid Esmaeili

unread,
Dec 15, 2015, 2:22:14 AM12/15/15
to Brendon Watson, Max Hunter, phy-users, klusta...@googlegroups.com
Dear Max,

Thanks for the valuable work.

I am trying to make sense of the probe files and make sure its what I need. As an example if we look at A1x32-7mm-100-1250-A32.prb, based on the defined geometry in this file it seems like if the probe sites are aligned from 1-32 along the shank, while it is not the case if we look at Neuronexus site mappings:http://neuronexus.com/images/Electrode%20Site%20Map/A32Maps_2013_11_22.pdf

Where  does the discrepancy come from? 

Best regards,
Vahid


--
You received this message because you are subscribed to the Google Groups "phy-users" group.
To unsubscribe from this group and stop receiving emails from it, send an email to phy-users+...@googlegroups.com.
To post to this group, send email to phy-...@googlegroups.com.
To view this discussion on the web, visit https://groups.google.com/d/msgid/phy-users/CAAqPF%2Be3LrBySQvMQbMr34Z-y5zro5p8LdwTvM8aaOXZER7TSw%40mail.gmail.com.

Justin Kiggins

unread,
Dec 16, 2015, 7:08:59 PM12/16/15
to phy-users
Vahid,

The discrepancy is as follows:

Phy needs the physical locations of the 0-indexed recording *channels*
The Neuronexus maps note the physical location of the 1-indexed recording *sites*

For numerous reasons, these will not match.
1. phy is 0-indexed while nn is 1-indexed
2. headstages/amplifiers/DACs may remap channels (that is, site 1 may correspond to channel 27 coming out of the headstage)
3. finally, software may remap channels (e.g. if channel 5 is bad and I choose to skip it when saving to phy's input format, then channels 6:n will shift down to be 6:n-1)

For all of these reasons, I went with the simplest strategy when generating these graphs: assume channels are saved from the tip, moving up the shank, from the leftmost to rightmost shank.

Doing this "right" (taking into account 1,2,&3 above) will necessitate a more flexible approach, such as using a dictionary that maps sites:channels. An example is here: http://www.justinkiggins.com/blog/generating-channel-graphs-for-klustakwik-automatically/ 

Best,
Justin

Vahid Esmaeili

unread,
Jan 11, 2016, 11:25:13 AM1/11/16
to Justin Kiggins, phy-users
Hi Justin,

Thanks for your reply, so as I was guessing one always needs a dictionary and those graphs are assuming that channels are starting from the tip.

Bests,
Vahid

Reply all
Reply to author
Forward
0 new messages