Syllable Annotation Guide
This guide presents a general protocol for annotating MoSeq syllables.
Considerations for Annotations
Try to annotate using a set of simplified and reusable tags that describe multiple instance of each syllable, including locomotive status, head and nose position, location within the arena, as well as higher-order descriptions. For rears, consider if the rear is wall-assisted or not. Examples: paused, scanning, sniff; or nose-down, dart; or rear-up; etc.
Annotating a Dataset
Load the MSQ file that you were tasked with annotating. For information on how to load your MSQ file, see the section Loading a Data File of the user guide.
Prepare your work area by adding any components, adjusting their size and position, and any component specific settings. You can alternatively load a MSL layout file (see the section Loading a Layout File). For convenience, we have prepared a MSL layout file which is appropriate for this protocol, and is available here.
Within your main filter(typically
filter0
):- Ensure that Count Method is set to Usage.
- Ensure that Selected Syllable is set to 0.
At this point, your screen should appear similar to the figure below:
Now, for each MoSeq syllable do the following:
- Select the current syllable by changing the Selected Syllable to the next syllable to be annotated. (If this is the first syllable to be annotated, you may leave this set to 0).
- Examine the Crowd Movies, Syllable Clips, and Spinogram components in order to garner an understanding of the behavioral ethology of the current syllable.
- Be sure to look at multiple individuals within the Crowd Movie. Reduce the playback speed if this is helpful.
- Be sure to look at all examples available in the Syllable Clips. Observe the videos with and without the Only Syllable Subclip option enabled. Reduce the playback speed if this is helpful.
- Be sure to look at all examples available in the Spinogram.
- Once satisfied with your impressions of the current syllable, place your annotations in the excel sheet.
- Go back to step 5.1 and repeat for all syllables.
Now, within your main filter (typically filter0), change the Count Method to Frames and change Selected Syllable to 0. Repeat step 5 while keeping the Count Method set to Frames until completed.