Here is a google doc of the living TrakEM2 alignment protocol. It is mostly complete but is being updated as we learn more each time.

UPDATE (2023-06-07):

If you are using a computer with many CPU cores, you need to make sure that you have enough memory (RAM) to handle your images for alignment. In general, you'll need 10x the file size of your largest 2D image for each core (see: https://imagej.net/plugins/trakem2/#how-much-ram-should-i-allocate-to-the-jvm-for-fiji-to-run-trakem2). EXAMPLE: If your dataset consists of images that are 600MB each and your computer has 12 cores, you will need at least 72 GB of RAM (= 600 MB x 10 x 12 cores).

FIJI/TrakEM2 at 3dem.org (Texas Advanced Computing Center) is deployed on stampede2, which has 48 CPU cores and 196 GB of RAM. It is currently configured to limit the number of cores that are available to 28, so that a tSEM image dataset (~600 MB per image) can be imported and aligned. When you start a DCV session from the 3dem.org's workbench, use SKX-normal or SKX-dev. If you have a dataset with larger images and need to adjust the number of cores, contact Masa or James.



Archival Protocols:

TEM Alignment in TrakEM2 (Cardona 2016) - PDF, Word doc.

NOTE: This is an outdated protocol for TEM that is being kept for archival purposes. Most of the relevant information for the Harris Lab has been transferred to the new TrakEM2 Alignment Protocol.