Covariance coloring on Rfam

Thanks Lauren (and Paul!) - this should help you along a lot!

Jim.

···

On 15/06/2010 00:31, Lauren Lui wrote:

Hi Jim,

I looked into the covariation coloring last night and asked Paul for
some more information about how they did it on Rfam. Below is his
response -- just wanted to keep you in the loop.

Best,

Lauren

---------- Forwarded message ----------
From: Paul P. Gardner<pg5@sanger.ac.uk>
Date: Mon, Jun 14, 2010 at 3:18 PM
Subject: Re: Covariance coloring on Rfam
To: Lauren Lui<llui@soe.ucsc.edu>
Cc: John Tate<jt6@sanger.ac.uk>

Hi Lauren,

Oooops, not sure how we lost that.

I've attached the missing doc. This should hopefully explain in gory
detail what we've tried to do.

I've also attached the code we use to generate these stats. This is
almost certainly not save for public viewing. Please don't laugh too
much. And there may be a few R18 comments. And well, good luck
deciphering that lot.

I haven't done much work optimising this measure. We do have a paper
where we try a whole bunch of different measures:
http://bioinformatics.oxfordjournals.org/cgi/content/abstract/22/24/2988

The covariation plots aren't something I look at very often. I tend to
find the sequence conservation plots a lot more informative. However,
I've never seen covariation marked up in an alignment in a useful way
so maybe this could improve things.

Good luck!
Paul.

On Mon, Jun 14, 2010 at 03:08, Lauren Lui<llui@soe.ucsc.edu> wrote:
   

Hi Paul,

I hope that you've been well. I'd like to add a coloring covariation
scheme to Jalview similar to what Rfam has. I'm looking at figure 2B
from the 2009 Rfam paper
(http://nar.oxfordjournals.org/cgi/content/full/37/suppl_1/D136/F2).

Can you provide me with more information about how this is calculated?
  The link to the description is broken
(http://rfam.sanger.ac.uk/family/colourings_desc.pdf). From the
references in the Rfam paper, I think that you are using Eq 4 from
this paper: Hofacker et al 2002 J Mol Bio
(http://dx.doi.org/10.1016/S0022-2836(02)00308-X), but I'd like your
confirmation on this. If it's possible to give me the code used to
generate the coloring that would be very helpful!

I might make a coloring scheme to apply to alignments, but I might
also create a covariation histogram below the alignment, similar to
the conservation histograms that Jalview already makes.

Thanks,

Lauren

--
  The Wellcome Trust Sanger Institute is operated by Genome Research
  Limited, a charity registered in England with number 1021457 and a
  company registered in England with number 2742969, whose registered
  office is 215 Euston Road, London, NW1 2BE.
     
--
-------------------------------------------------------------------
J. B. Procter (JALVIEW/ENFIN) Barton Bioinformatics Research Group
Phone/Fax:+44(0)1382 388734/345764 http://www.compbio.dundee.ac.uk
The University of Dundee is a Scottish Registered Charity, No. SC015096.