Installation on macOS just for analysis

[marsiwiec]

Hi,

I was wondering if it is possible to use the MIES module psx analysis (new feature in 2.7) for postsynaptic current detection on macOS, without hardware XOPs. I watched the YouTube video and it looks pretty great. I tried setting it up according to manual installation instructions but there were errors on startup.

Thanks,
Marcin

[t-b]

@marsiwiec Can you post the error message?

I bet it is due to the missing XOPs for MacOSX.

[t-b]

A followup question: You are using MIES to acquire your data as well or? Or who have you planned to get your data into MIES?

[marsiwiec]

Right now I am using ACQ4 on a Sensapex rig (scope + manipulators + pressure controller), and I am really enjoying the semiautomatic patching capabilities. As an aside, do you have any workflows for combining approaching/patching cells with ACQ4 and subsequent recordings with MIES? I've noticed some scripts inside the ACQ4 repo but cannot figure out how to combine the two.

I would like to switch to MIES for executing certain recording protocols as I find it a lot more responsive and powerful especially for multipatch. The recordings are performed on a Windows machine, so the macOS MIES would only be involved in occasional PSC analysis.

I am also trying to import ABF files into MIES to analyse postsynaptic currents. Do you think a simple conversion to nwb2 using x-to-nwb or similar would be enough to be able to do the analysis?

[t-b]

Right now I am using ACQ4 on a Sensapex rig (scope + manipulators + pressure controller), and I am really enjoying the semiautomatic patching capabilities. As an aside, do you have any workflows for combining approaching/patching cells with ACQ4 and subsequent recordings with MIES? [..]

@timjarsky will follow up on that one.

I would like to switch to MIES for executing certain recording protocols as I find it a lot more responsive and powerful especially for multipatch. The recordings are performed on a Windows machine, so the macOS MIES would only be involved in occasional PSC analysis.

Nice to hear that you like MIES. I've created #1639 to track the addition of experimental MacOSX support without hardware.

I am also trying to import ABF files into MIES to analyse postsynaptic currents. Do you think a simple conversion to nwb2 using x-to-nwb or similar would be enough to be able to do the analysis?

At the moment I don't think this is possible. I've created #1858 to track that.

[campagnola]

@marsiwiec the MIES bridge has only been used with an older version of ACQ4 (much older; python 2 only) and does not work with the latest. We will be working on this over the next several months, though, so stay tuned.

[marsiwiec]

@campagnola thanks for the info, I will stay tuned :)

[t-b]

@marsiwiec In #1859 I have prepared a first version of MIES with experimental MacOSX support. It passes the relevant part of the non-hardware tests and also includes all XOPs required for analysis. Can you give that a shot and report back if it works on your end?

[timjarsky]

@marsiwiec I've contacted @campagnola; he should be in touch soon about the ACQ4 code for MIES-ACQ4 integration (That's how we use it at the Allen Institute)

[marsiwiec]

@t-b Thanks! I will give this a try soon and report back.

@timjarsky Thank you. I look forward to integrating these tools. One more question - is there a way to analyze ACQ4 files in MIES? I have multisweep recordings of sIPSCs/mIPSCs done in ACQ4 and would love to analyze them in the new psc analysis module.

[t-b]

One more question - is there a way to analyze ACQ4 files in MIES?

Not that I know of. I've created #1869 to track the request.

[marsiwiec]

Thanks. Maybe "dressing up" the ACQ4 dataset and exporting to NWB somehow would make it compatible? I honestly have no clear idea how to achieve that though. I think I will just try to use MIES for recording PSCs from now on. Planning to manually switch from patching a cell in ACQ4 (to make use of the semiautomatic capabilities) to recording protocols in MIES, as it is definitely superior in terms of data acquisition/analysis.

One more question - I am trying to set everything up and it would really help if I had a sample config file for one of your rigs, just to see how to set everything up. Would that be ok?

[t-b]

I don't have access to the rig config files, but we do have some shipped with MIES, see https://github.com/AllenInstitute/MIES/tree/main/Packages/Settings_template.

[marsiwiec]

@t-b so I cloned the repo, switched to the 1859 branch, symlinked everything according to the documentation and although Igor starts I get this message in the console:

Experiment: Untitled () Igor Pro version: 9.02 (39951) ################################ Double threadsafe assertion Fail encountered !

Also, there is no menu item for MIES.

[t-b]

Okay, that was helpful. I've pushed another fix, please update to 9f35ff1 (GetDrive: Make it work on Macintosh, 2023-10-25) on that branch.

[marsiwiec]

Ok, thanks, now it starts with no errors! Is there anything in particular you would like me to test for? I do not have sample traces recorded with MIES with me now, but might be able to get something next week.

[t-b]

Ok, thanks, now it starts with no errors!

Very nice!

Is there anything in particular you would like me to test for? I do not have sample traces recorded with MIES with me now, but might be able to get something next week.

I would not expect any big issues now that we got it so far. Do I'd say we merge the PR and you can then followup if something else comes up?

[marsiwiec]

I would not expect any big issues now that we got it so far. Do I'd say we merge the PR and you can then followup if something else comes up?

Sounds great. I plan to start incorporating MIES into my daily recordings and gradually test the macOS analysis. Thanks!