annotation for nucleotide alignments

Hi there,

I'm trying to color a nucleotide alignment by annotation (conservation in this case), so that only the non conserved nucleotides are highlighted in color. See attached document below for an idea of what I need. I managed to do it previously (I was working with a previous version of Jalview - v9 I believe for macs - but I am not sure the issue comes from that).
Now, I'm trying to do the same with another alignment and with Jalview v10, but it is not possible to select the option "by conservation" (and the accompanying "Modify Conservation Threshold..."), and when I select the option "by annotation", the window that pops up is empty. Coherently, when I select "Show annotations", it says that the alignment has no annotations.
So how can I transfer the conservation plot that is visible in ClustalX for instance (grey bar with dents when nucleotides are not conserved) to annotate my alignment in jalview? Or should I save the alignment under a different format than the clustal format (.aln) to have this conservation included in the file? Note: I usually use muscle to align sequences.
Sorry for this very basic question, and thanks in advance for your help!
Best,

Steph.

aln_small.pdf (8.75 KB)

Hi Stephanie - thanks for resending your email.

I'm trying to color a nucleotide alignment by annotation (conservation in this case), so that only the non conserved nucleotides are highlighted in color. See attached document below for an idea of what I need.

understood.

I managed to do it previously (I was working with a previous version of Jalview - v9 I believe for macs - but I am not sure the issue comes from that).
Now, I'm trying to do the same with another alignment and with Jalview v10,

whoa.. can I ask you where you have got these version numbers from ? The current version of the application available from http://www.jalview.org/download.html is 2.4.0b2 - and assuming that you are using the Jalview Desktop application, you should be able to see this in the 'About' dialog box available from the Desktop window's Help menu.

but it is not possible to select the option "by conservation" (and the accompanying "Modify Conservation Threshold..."), and when I select the option "by annotation", the window that pops up is empty.

There are a number of things that may cause this:

1. Make sure you have the 'caculate consensus' preference enabled - otherwise Jalview will not automatically calculate any consensus annotation for the alignment. To do this, open the 'preferences' dialog (under the Desktop's Tools menu) and check that the box next to 'consensus' is ticked (if it isn't, tick it, click ok, and then reload your alignment).

2. are you working with a very big alignment ? if so, then you may have run out of memory - jalview will give up trying to calculate the consensus histogram if there are too many sequences. Check the 'Show memory usage' box under 'Tools' to see if you are low on memory. - and have a look at the faq entry to see how to increase it if you have less than 50% left with just one alignment loaded ( http://www.jalview.org/faq.html#incmemory ).

3. there is a remote possibility that you are using one of the mac java versions which Jalview doesn't play well with. Check the java settings on your machine to make sure you are using a java 1.6 virtual machine (see this post, for instance: http://www.compbio.dundee.ac.uk/pipermail/jalview-discuss/2009-March/000164.html ).

I'm hoping that #1 provides the answer to this problem, or if not - #2 is the cause. #3 is an outside bet, but you would be having many other problems if you are using an incompatible JVM. Please let me know if none of these fix the problem!

best of luck..
Jim.

···

On 18/03/2010 16:26, Stephanie Blandin wrote:

--
-------------------------------------------------------------------
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.

Hi Stephanie.

Thanks very much for your quick answer!
   

no probs.

Well, sorry about that. It is indeed the latest Desktop version I am using on a Mac (OS 10.6.2). When I open Jalview, it says 2.4.0.b2, but if I choose "About Jalview" in the Jalview menu next to the apple on the upper menu bar, it says v10 :wink:
   

ah. hmm. I'll have to take a close look at where that comes from. I don't remember seeing that on my own 10.6 install.

The "autocalculate consensus" was indeed enabled. The "conservation" box is also checked in preferences/visual/open new alignment, and actually, it works very well when I open an amino acid alignment: I do see the conservation annotation when working with aa aln, and I can color my alignment by annotation/conservation then.
   

That's good. I should say here that the reason you don't see a conservation trace for the nucleotide alignment is because the conservation calculation is based on amino acid chemical property conservation - so its disabled when working with nucleotide alignments. However, you should still see the consensus trace (and be able to use it via the 'colour by annotation' dialog). One thing that affects the consensus calculation is whether there are lage deletions in many of the sequences - if this is the case with your particular alignment then right click on the consensus row to enable the 'ignore gaps' option.

I checked this and it is fine. The specific aln I was working with include 2-6 sequences, up to 5kb or so. There is more than 70% of free memory.
   

ok. definitely no problems there at all.

3. there is a remote possibility that you are using one of the mac java
versions which Jalview doesn't play well with. Check the java settings
on your machine to make sure you are using a java 1.6 virtual machine
(see this post, for instance:
http://www.compbio.dundee.ac.uk/pipermail/jalview-discuss/2009-March/000164.html
).
     

The Java version I have is 1.6.0_17 from Apple Inc. (it includes a 32-bit and a 64-bit version of Java SE6). And the rest of the tools/window sizing/etc that I have tested work well.
   

again, no problems.

I could try to open my aln on a different computer equipped with an older version of Jalview. I'll keep you updated.
   

that might help you - but I'd like to get to the bottom of this since it looks like the 2.4.0b2 version of jalview is working fine - but apparently not letting you do what you want. If the above comment about the consensus annotation doesn't help then would it be ok if you sent me an alignment that demonstrates the problem ?

Jim.

···

On 19/03/2010 10:17, Stephanie Blandin wrote:

--
-------------------------------------------------------------------
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.

Dear Stephanie.

Good to hear the issue is cleared up!

Excellent, and sorry I did not think of that earlier...

its understandable. This information is buried in the documentation, and that seems to be the problem !

This time, I also enabled the "consensus" option, so the consensus
appeared in the annotations and I was able to color the non-conserved
nucleotides by coloring "by annotation / consensus / below threshold".
So this solved the problem, and I guess I was doing it like that
previously, as it did not work either when I tried to color the aln by
conservation using the previous version of Jalview. I will make a note
of it, not to forget the trick again :wink:

Hopefully, there will be more default preferences in the future so you can simply set the default to 'colour not 100% conserved regions' - I'm thinking this will be something like an 'alignment visualization stylesheet' that you can configure yourself.

Actually, I must also confess I tried to trick the program by
including amino acid letters at the beginning of the sequences (by the
way, is there a way to declare it is a nucleotide or an amino acid
alignment?), but it did not work either, Jalview still considered my
sequences as nucleotides, and ignored the few aa at the beginning of
the sequences...

jalview picks the type by working out the proportion of ACTGU symbols in the alignment - we did consider allowing the type to be changed, but it wasn't really a priority to implement it. I'm sure it'll happen at some point, but the more interesting things would be to make it nicer to work with nucleic acid sequecnes in Jalview - so - for instance - extending the conservation analysis so it calculates conservation for the predicted codon translation at each position in a nucleotide alignment, rather than simply disabling it.

I'm happy to send you the aln if you still want to have a look at it,
just let me know.

Thanks very much for your help!

a pleasure... Happy Jalviewing!
Jim.

···

On 22/03/2010 10:44, Stephanie Blandin wrote:

--
-------------------------------------------------------------------
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.