voxel_vote: command not found, disk quota, overlap issue

[gk1250]

Hi, I have some questions about the errors in running the magetbrain.

salloc --x11 -p day -t 1- -c 36 --mem-per-cpu 20g source MAGeTbrain/bin/activate module load parallel module load Python/3.10.8-GCCcore-12.2.0 source MINC2/minc-toolkit-config.sh export QBATCH_PPJ=8 export QBATCH_CHUNKSIZE=1 export QBATCH_CORES=1 export QBATCH_MEM=32G mb status mb run --stage-templatelib-procs 1 --stage-voting-procs 8

These are the commands I used to run the magetbrain.

I was wondering if "/usr/bin/sh: voxel_vote: command not found" is causing the disk quota to exceed. I keep getting 500+ mini files in the output/intermediate directory (outside the template folder, nothing in the fusion folder) and I'm not sure it's normal. If it's an error, could you tell me how to fix it? Would incorporating files you had uploaded in the antsRegistration-MAGeT/bin folder such as mb_vote.sh be helpful in fixing this issue?

Also, I manually did AC-PC alignment for each subject's T1 MPRAGE image (RAS oriented, skull stripped, no smoothing, no co-registration), and I got this error message: MattesMutualInformationImageToImageMetrics4(0x2c97790): Too many samples map outside moving image buffer. There are only 7 valid points out of 21384 total points. The images do not sufficiently overlap. They need to be initialized to have more overlap before this metric will work. For instance, you can align the image centers by translation. Would ignoring this error message be okay?

Thank you so much!

[gdevenyi]

You have not installed all the dependencies.

https://github.com/cobralab/MAGeTbrain#dependencies

As for alignment, you should be performing no prealignment at all, simply inputting native-space images. Performance is better with preprocessing and skull-stripping but not required.

I suggest you take a look at https://github.com/cobralab/MAGeTbrain-nextflow which implements a nextflow implementation with containerized runtime and supports nifti files for all inputs. I'll try and get around to adding some basic instructions, you can take a look at the nextflow.

Finally, you have not described at all the input configuration you are using, the number of files, the atlases, labels, their origins, etc. PLease do so for any further support requests.

[gk1250]

Hi Dr.Devenyi,

Thank you for your email.

I'm using 5 habenula segments that are publicly available on the CoBrALab GitHub page for the atlases and 50 children's T1 MPRAGE data for the subjects (I selected 21 of them with the lowest average motion for the templates).

As you suggested, I ran the program again with skull-stripped data and no prealignment, but got the same error message. I got this error message for 18 participants' data, and somehow they are all part of the templates folder. [image: overlap error.png] I was wondering if there are ways to handle this error or if it would be okay to ignore it and continue with the voting process.

Thank you so much for your help.

Best regards, Gihyun

ᐧ

On Thu, Mar 14, 2024 at 11:18 PM Gabriel A. Devenyi < @.***> wrote:

You have not installed all the dependencies.

https://github.com/cobralab/MAGeTbrain#dependencies

As for alignment, you should be performing no prealignment at all, simply inputting native-space images. Performance is better with preprocessing and skull-stripping but not required.

I suggest you take a look at https://github.com/cobralab/MAGeTbrain-nextflow which implements a nextflow implementation with containerized runtime and supports nifti files for all inputs. I'll try and get around to adding some basic instructions, you can take a look at the nextflow.

Finally, you have not described at all the input configuration you are using, the number of files, the atlases, labels, their origins, etc. PLease do so for any further support requests.

— Reply to this email directly, view it on GitHub https://github.com/CoBrALab/MAGeTbrain/issues/61#issuecomment-1998866714, or unsubscribe https://github.com/notifications/unsubscribe-auth/BGHAYDH34FHPMN2LVYDAZBTYYJR7FAVCNFSM6AAAAABEXHCMHSVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMYTSOJYHA3DMNZRGQ . You are receiving this because you authored the thread.Message ID: @.***>

[gdevenyi]

You have not installed all the dependencies. https://github.com/cobralab/MAGeTbrain#dependencies

[gk1250]

Hi, thank you for your email.

I emailed the HPC to install the necessary modules on the cluster. I'll let you know if I still have the same issues.

Best regards, Gihyun

On Mon, Mar 18, 2024, 10:06 AM Gabriel A. Devenyi @.***> wrote:

You have not installed all the dependencies. https://github.com/cobralab/MAGeTbrain#dependencies

— Reply to this email directly, view it on GitHub https://github.com/CoBrALab/MAGeTbrain/issues/61#issuecomment-2004015741, or unsubscribe https://github.com/notifications/unsubscribe-auth/BGHAYDHNK72TTXJOBNG7CWLYY3YG3AVCNFSM6AAAAABEXHCMHSVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZDAMBUGAYTKNZUGE . You are receiving this because you authored the thread.Message ID: @.***>