This morning I stumbled upon the ChromoZoom genome browser and I must say I am very impressed. This browser seems to be amazingly fast and its operation is surprisingly smooth! What kind of twilight zone is this - where a graphical software is as easy to use at the authors claim it to be ...
There are a few small usability issues: it was not immediately obvious to me that the way to resize the tracks is to drag on separator marker on the left side, or that the scroll wheel zooms in an out. But other than that it really flies!
More on how it works: it actually serves image tiles scraped from the UCSC genome browser, so we can think of it as a Google Maps like software. IMHO the two aspects, tile scraping and browser interface should be separated with a clear specification
Does look nice to use, but IIUC, it is limited to a small subset of the tracks available from UCSC, no?
It is nice that is lets you load your own tracks though.
Correct, the current tracks we have take up about 64GB on disk, which we could probably improve with some trickier compression tricks, but we're not sure which of the additional tracks people actually use enough to justify the cost.
We've considered converting them all to bigBed and bigWig and letting people add them that way, though. Performance for those two formats is really good and the code for rendering them is surprisingly simple, with room for all kinds of drawing extensions. There's a lot you can do in <canvas> and its performance is pretty good now!
Hey Istvan, I fixed your issue with the need to introduce new users to the interactions. Would love if you could pop open a fresh browser with no cookies (or private/incognito mode, etc.) to simulate a new user and see if it works for you when you visit http://chromozoom.org . Cheers!
Hi, author of ChromoZoom here! It's a pleasure to hear that people are using the software--I believe it is being used for some courses on sequencing at the U of Toronto, and among some of the yeast labs there that pump out lots of MiSeq data.
Yep, I am a big believer in usability, I worked on web front-end code for a few years before moving back into bioinformatics. Also, a big fan of Edward Tufte, hence the minimalist interface and tick marks.
I thought that I had the "Getting started" box open automatically for new users, which explains the exact two interactions that confused you, but something must be funky with my cookie detection code. I've put it on my bug tracker. Edit: this issue is now fixed.
Does look nice to use, but IIUC, it is limited to a small subset of the tracks available from UCSC, no? It is nice that is lets you load your own tracks though.
yes, I think this is how much they scraped off at this moment - scraping all tracks and storing them locally as images would be quite a job
Correct, the current tracks we have take up about 64GB on disk, which we could probably improve with some trickier compression tricks, but we're not sure which of the additional tracks people actually use enough to justify the cost.
We've considered converting them all to bigBed and bigWig and letting people add them that way, though. Performance for those two formats is really good and the code for rendering them is surprisingly simple, with room for all kinds of drawing extensions. There's a lot you can do in
<canvas>and its performance is pretty good now!Hey Istvan, I fixed your issue with the need to introduce new users to the interactions. Would love if you could pop open a fresh browser with no cookies (or private/incognito mode, etc.) to simulate a new user and see if it works for you when you visit http://chromozoom.org . Cheers!
nice animations! Very impressive tool!