Looking for recommendations on structuring a track hub

[Melissa Cline]

Hi there!

I'm working on a track hub for MaveDB, which is a repository of functional assay scores.  I prototyped the track hub as a bigwig (see attached), but I'm not sure that's the best approach.  I'd like a recommendation on what type of track to use, given the data:

• There are one or more functional assays reported.  Each assay covers a discrete piece of the genome, less than one gene.  When we scale up, there will be thousands of functional assays
• Each functional assay reports the functional impact of N mutations.  For each mutation, the reported data includes the reference base (or amino acid), the alternate base / amino acid, and a numeric score.  There can be multiple mutations per genomic base.

I prototyped the track hub for a single assay as a bigwig, as shown, but it omits a lot of information.  I randomly selected 1mutation per base.  The bar height shows the score, but what's not shown is the mutation.  The ClinVar track in my screenshot seems like a better example, with one entry per mutation per base, but I'm not sure about the best way to represent the score.  For track users, the exact numeric score probably isn't as important as a relative score (i.e. high/low/moderate for the assay), so perhaps a colored block could work, esp if the colors can be scaled dynamically rather than discretely.  Eventually, we can link back to MaveDB itself for anyone who wants to drill down into the actual values.

Thanks!

Melissa