Importing skeletonised neurons into NAT

[Joses Ho]

This seems like a basic question, but I can't seem to piece together from the documentation how to move from the skeletonised neuron data (which ImageJ outputs as csv data) to the RDA or SWC format favoured by the 'nat' package. Any little bit of help will be much appreciated!!

Gratitude in advance,

Joses

[Gregory Jefferis]

If you are using Fiji's simple neurite tracer plugin then you have the option to export as SWC (simplest) or in an XML format with extension .traces which is the plugin's native format. 

Bay's read.neuron function understands both of these formats.

There is also a csv option but this loses all connectivity information and there is no direct support in read.neuron for this reason. 

Best,

Greg. 

Sent from my iPhone

--
You received this message because you are subscribed to the Google Groups "nat-user" group.
To unsubscribe from this group and stop receiving emails from it, send an email to nat-user+u...@googlegroups.com.
Visit this group at http://groups.google.com/group/nat-user.
To view this discussion on the web, visit https://groups.google.com/d/msgid/nat-user/93b596a4-2033-44be-9427-5f4892488cd1%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[Joses Ho]

Thanks for this (works perfectly if I am using the Simple Neurite Tracer plugin). 

However now I am interested in importing the results of the AnalyzeSkeleton plugin (is CSV format) into NAT. Is this possible?

Thanks,

Joses

[Gregory Jefferis]

Dear Joses,

> On 13 Aug 2015, at 04:40, Joses Ho <jose...@gmail.com> wrote:
>
> However now I am interested in importing the results of the AnalyzeSkeleton plugin (is CSV format) into NAT. Is this possible?

One option is to import the image as generated by Skeletonize 3d (the step before AnalyzeSkeleton) as a dotprops object. This will be perfect for NBLAST and also works fine for visualisation. To do this,

1. from Fiji save your skeletonised image as a Nrrd file
2. Be sure to hold down the Alt (or Option) key to save in compressed format (which will be tiny)
3. From R call dotprops with the name of your nrrd file.

Two tweaks I would recommend.

1. Downsample the input image to roughly 1um isotropic voxels before skeletonisation.
2. Add k=5 as an extra argument to the call to dotprops to reduce the number of nearest neighbours used to calculate the tangent vectors.

Finally if you do want to import the csv format as either a fully connected tree (neuron object) or a vector cloud (dotprops object), that will not be hard, but please post a sample csv file eg as a GitHub gist and I can respond with a code snippet.

Best,

Greg.

[Joses Ho]

Will give your suggestions a go! Thanks a lot.

Regards,

Joses