What interactive/3D visualization tools do bioinformaticians wish existed but don't?

Question

[joseph.thomas.heenan]

What are some of the most difficult visualization challenges bioinformaticians face today? Which 3D visualization tools do folks spend the most hours using? Which tools do you wish existed? Is it visually inspecting protein structures / PDBs in a tool like iview or jmol for homology detection or protein/ligand binding a common or important task? Do folks here spend time viewing de novo assembly graphs in a tool like Bandage? What other use cases might I be missing?

Background: I am a CS Grad Student with experience in augmented reality development for HoloLens and would like to develop some open source bioinformatics visualizations tools as part of my studies. Any ideas/suggestions/pain points you can share would be much appreciated. I am thinking that being able to visualize a protein as a hologram and collaborate on it together with colleagues could be interesting, however I could really use some more help from the professionals here on use cases (I'm not really sure what "collaboration" would even mean - maybe picking up a complementary ligand hologram and checking fit?). Thanks!

Joe

Answer 1

[Alex Reynolds]

Being able to visualize long-range chromatin interactions with Oculus or similar VR gear would be interesting, to align tertiary DNA structure with methylation, variation, Hi-C, and other signals, especially where structure and signal change in different tissues, drug treatment conditions, editing with CRISPR/Cas9 or TALENs, etc. There are tools to explore these datasets on 2D computer screens, but this kind of data seems ripe for a true 3D approach that invites grabbing different regions and doing interactive exploration. Maybe you could collaborate with these folks and get their Oculus support going.

[joseph.thomas.heenan]

Alex - When you say "There are tools to do explore these datasets on 2D computer screens" - what type of tools could I look at as a reference? Thanks in advance

[Alex Reynolds]

Tools might include WashU, UCSC, and Hi-C genome browsers, which use a 2D stack of tracks to visualize signal.

Answer 2

[Jean-Karim Heriche]

Not strictly bioinformatics but a problem that would benefit from real 3D visualization is tracking of objects in time-lapse 3D microscopy images.

[joseph.thomas.heenan]

Thanks Jean-Karim! Do you know of any database that has these time-lapse images? Would they be in PDB format?

[Jean-Karim Heriche]

I don't think there's a database with these at the moment. Anyway, the format would not be PDB since PDB is designed for macromolecular structures. Images are unfortunately represented in a wide variety of formats. As an example of the problem I am referring to, have a look at this paper.