A Group is a struct that stores a list of (match) sequences, along with any properties that are relevant (e.g. highlight color, visibility etc.). A Selection is essentially just a list of sequences too, so many functions are designed to operate on a list of sequences so that they work with either a Group or a Selection. The Groups functionality is very basic at the moment, offering just a simple mechanism to highlight or hide a set of sequences, but it could be extended to give much more powerful filtering and sorting options.
</p>
A Group is a struct that stores a list of features, along with any properties that are relevant (e.g. highlight color, visibility etc.). A Selection is essentially just a list of features too, so many functions are designed to operate on a list of features so that they work with either a Group or a Selection. A Filter is also just a Group with the "filter" flag set. Filtering works as follows: if there are no groups with the filter flag set, then all features are shown; if there is one or more group with a filter flag set, then only features in that group (or another group with the filter flag set) are shown. Other features are filtered out, but only if they are the same type as the features in the Filter group(s). This is because generally an annotator wants to filter out sequences to only see a particular sequence, but they still want to see transcripts and other feature types. This has evolved over time so is not as intuitive as it could be when you start creating more than one group/filter. However, most annotators just want to quickly create one filter, in which case it works fine from a user's perspective.
</p>
<p>
Groups can be created from the clipboard (i.e. current selection in blixem or zmap) or from a search. The functionality could be expanded to create groups in other ways, e.g. based on %ID.
