We are pleased to present the release of v1.7.5
This release includes two major improvements
- Ability to index tracks for searching by gene names in JBrowse Desktop. This let's users choose per-track indexes in the add track widget interactively, or index them after-the-fact using the track menu.
- Removed serialization of BAM/CRAM features for pileup tracks which brings large performance benefits for jbrowse-web and jbrowse-desktop. Previously, all the BAM/CRAM features were serialized across the RPC boundary in jbrowse-web and jbrowse-desktop, but this was an expensive and memory intensive operation. Now, the BAM/CRAM features are not serialized, but the main thread has to manually request any data it needs from the worker after rendering (e.g. when we click a feature).
Figure showing the "Index track" option in the track menu of JBrowse Desktop
Figure showing the "Index track" option in the add track widget of JBrowse Desktop
Figure showing performance improvement of the noserialize improvement compared with v1.7.4, which brings the webworker performance in line with the embedded app (which does no serialization)
This release also fixes a bug with using trix indexes created by
jbrowse text-index in v1.7.0-v1.7.4 (with the adjustable prefix size)
To install JBrowse 2 for the web, you can download the link above, or you can use the JBrowse CLI to automatically download the latest version. See the JBrowse web quick start for more details.
See "Packages in this release" in the changelog for links to packages published on NPM.
- #2684 Text-indexing in desktop (@teresam856)
🐛 Bug Fix
- #2863 Render gene with CDS subfeatures properly (@cmdcolin)
- #2934 Bump @gmod/trix to fix prefix size calculation and searching first word in index (@cmdcolin)
- Colin Diesh (@cmdcolin)
- Garrett Stevens (@garrettjstevens)
- Teresa Martinez (@teresam856)