Skip to main content

2024 year in review

· 5 min read

Hello all,

I know we are well into 2025 already, but I wanted to review some of the great progress JBrowse made in 2024!

Without further ado, here are some highlights

New multi-sample VCF displays​

To help render population-scale data, we created new "multi-variant" rendering modes. This includes a "matrix" rendering mode that is a dense visualization of variation patterns.

This was an exciting development that has created a flurry of feature development

One particularly interesting feature is the "phased rendering mode", which can split the phased genotypes into separate rows, which effectively shows the individual haplotypes.

The "phased rendering mode" has the unique ability to help spot patterns of inheritance in e.g. trio datasets. You can visually see where a child inherited particular variants from a particular parent

The data in the multi-variant display can be sorted according to sample metadata, which can be imported from a "samples TSV", so users can group and color the samples by e.g. the population code (here showing 1000 genomes data)

Multi-way synteny views​

We created the basis for rendering multi-way synteny views. Currently, this is based on "series of pairwise" comparison rendering, which has certain drawbacks (it is not a true multi-way comparison) but it is a big stepping stone

We intend to create more improvements allowing users to load multi-way comparisons from a single data file, and easier UI workflows for setting up multi-way views in the future

Inversion SVs enhancements for breakpoint split view​

We created new enhancements to the breakpoint split view to allow it to work better in a 'single row' configuration, and to color reads by pair orientation or by changes in direction in a split alignment

Screenshot showing read pairs being connected from a variety of sequencing runs from Illumina, PacBio, and Nanopore from the Genome in a Bottle dataset

Note: There were many other improvements to the breakpoint split view including, importantly, the ability to launch it from non-SV-inspector workflows. Previously, the SV inspector was the only way to really launch a breakpoint split view, but now it can be launched from just clicking the feature details of a read or VCF feature.

New "SV inspector" improvements​

  • Allows opening tracks from the config.json in the SV inspector and spreadsheet
  • Uses @mui/x-data-grid for display of the spreadsheet
  • Makes separate columns for VCF INFO fields
  • Support larger files by storing data in 'volatile' storage that doesn't bog down the session
  • Support DEL, INV, and all types of SVs instead of just TRA and BND type variants in the SV inspector
  • Re-navigates existing breakpoint split view instead of launching a new one each time

Screenshot showing the new SV inspector user interface

New "Recent sessions" menu with autosaves​

We saw users have trouble where they would close their tab containing JBrowse 2 and they would lose their work.

It would be difficult to restore where they were at

To help this situation, we created a new system where sessions are AUTOSAVED! Yes, autosaved! So if you accidentally close a tab or come back the next day, it is easy for you to go back and restore where you were at:

Screenshot showing the ability to access "File->Recent sessions"

  • You can also create "Favorite" sessions to revisit for later
  • Admins can also create "Pre-configured" sessions for their users

If you have any feedback about this feature let us know. I think the discoverability of this feature could use more improvement

New "Group by" functionality for alignments tracks​

We released new "Group by" functionality for the alignments tracks. This functionality is somewhat unique because applying it creates multiple functionally independent subtracks.

The effect of being split into independent subtracks is actually different from e.g. the Group by feature in IGV, because in IGV, only the "read stack" is split into different rows, whereas with JBrowse, each group gets its own read stack and coverage calculation

image8

Example with splitting an alignments track by the HP tag

Apollo 3 beta release​

Apollo 3, the genome annotation editor built on JBrowse 2, went into public beta at the end of 2024!

See release announcement here https://github.com/GMOD/Apollo3/releases/tag/v0.3.0

The Apollo 3 editor can run as a 'collaboration server' (shared annotation editing on a website) or as a standalone JBrowse 2 desktop plugin. It is able to take full advantage of JBrowse 2 capabilities, and users can annotate e.g. inside of a synteny view

image

Feel free to get in touch for more information on the project!

The year ahead​

We look forward to hear from more users. Our office hours outreach effort has been very successful with over 100 one-on-one meetings with users! Feel free to sign up for one at https://jbrowse.org/jb2/contact :)