What sequence browsing tools have players developed?

  • 2
  • Question
  • Updated 5 years ago
  • Answered
Our lab is working on a better browser and perhaps API for the RNA Mapping Database (RMDB), hosted at Stanford.

As many of you know, the RMDB has served as the primary repository of EteRNA SHAPE data as well as other projects, and its got a pretty good browser but it could be improved.

I know that the player community has generated their own amazing tools for browsing & searching these data -- could you possibly put links here to your currently favored tools?

We'd like to review them to see what was useful, and perhaps even deploy features created by you in the RMDB (giving credit of course). (We might even use this as a basis for revising the EteRNA in-game sequence browser, which I know can be improved.)
Photo of rhiju

rhiju, Researcher

  • 404 Posts
  • 123 Reply Likes
  • excited

Posted 5 years ago

  • 2
Photo of jnicol

jnicol, Player developer

  • 57 Posts
  • 21 Reply Likes
Eli has several analysis tools listed on his profile page, http://eterna.cmu.edu/web/player/8627/ 
Search for Lab Data Analysis tools. You may need to contact the authors to verify their latest version.
Photo of Eli Fisker

Eli Fisker

  • 2275 Posts
  • 517 Reply Likes
My absolute favorite lab analysis tool is jandersonlee's LabDataMiner. This tool is brilliant.

http://eternawiki.org/wiki/index.php5...

Plus my favorite introduction to it:
https://docs.google.com/document/d/1S...

Other things that will be particular useful, is what was made in combined effort of Meechl and Omei.

http://eternawiki.org/wiki/index.php5...

Meechl put all the RMDB resources in spreadsheets and Omei put them in a fusion table. The strength of the latter is that is is easier to search across labs.

Tools like fusion tables and spreadsheets allow more players to make statistics and graphs.

It would be amazing having the option to make statistics or graphs about things like which base was present at what position in an RNA sequence.

Meechl has made beautiful graphs and I would very much like a took that would make it easy to make graphs of specific features. To my knowledge she has used R to make the graphs.

A few examples of her in action:

https://getsatisfaction.com/eternagam...

https://getsatisfaction.com/eternagam...
Photo of Eli Fisker

Eli Fisker

  • 2275 Posts
  • 517 Reply Likes
New awesome fusion table by Omei:

Fusion Table for MS2 Riboswitches on Chip (Round 1)

Details can be found here:
https://getsatisfaction.com/eternagam...
Photo of Omei Turnbull

Omei Turnbull, Player Developer

  • 1000 Posts
  • 322 Reply Likes
Thanks, Eli!

Rhiju, the main reason I originally got interested in loading Eterna data into fusion tables was for the API they provided for more specialized, RNA specific, tools. Coding tools to the Eterna API has been an exercise in frustrating for a number of reasons:

1) The Eterna database APIs are largely tailored to the individual Eterna Web pages. If a tool wants to do something different, it often has to make many individual queries to gather the data it needs.
2) Because of point 1, as the Eterna pages have changed, the database queries to support them have changed, meaning that 3rd party tools become obsolete. Since there has been essentially no regular communication between inside and player developers, this discourages player development.
3) The Eterna API doesn't support JSONP or other more modern standard that allows Web application pages hosted on non-Eterna servers to query the Eterna database servers without going through an intermediate proxy like YQL.

Putting Eterna data into fusion tables was something I could do on my own that would fix all these issues. Fusion tables support JSONP and automatically provide a very orthogonal SQL-like query interface so that developer tools aren't continually being broken by wandering APIs. The fact that fusion table also provide a serviceable UI was a somewhat unexpected bonus, because so many useful things can actually be done without an RNA-specific browser.

I bring this up here because I would suggest that if you are going to improve the accessibility of RMDB data, putting effort into a good general-purpose API is at least as important as writing a better browser application. Not only will your browser be easier to enhance and evolve over time, but users will be much more empowered to make their own improvements.
Photo of rhiju

rhiju, Researcher

  • 404 Posts
  • 123 Reply Likes
Outstanding feedback.
We'll be revisiting sequence browsing and APIs in the context of RMDB in February.

It sounds like at that point we should propose a tentative API and ask for specific comments form the community before implementation.
Photo of jandersonlee

jandersonlee

  • 555 Posts
  • 130 Reply Likes
I'd love to give feedback on an API or to participate in the initial design.
Photo of jandersonlee

jandersonlee

  • 555 Posts
  • 130 Reply Likes
In addition to the LabDataMiner Eli mentioned above, I've played with a DotPlot tool that lets you explore the dot-plot graph and the underlying RNAsubopt space (alternative foldings) for a given sequence. Not quite the same as the search tool.

I also have some rough working notes for a CASE (Computationally Assisted Selected Elements) tool that will let you break down a target secondary structure into component elements with overlapping constraints and then search a database for possibly matching components. Having a standard search API for an RNA database might make this even more useful. I could foresee for example having some of Cody's motifs available for loops etc and searching lab results and other databases for matching stems/stacks. This is still very much a thought experiment in progress more than a work in progress.