Problem with saving edits in CONSED 19.0

[nbenzakour]

Hi everyone,

I'm currently working on an hybrid assembly Sanger/454 using Consed as an editor and encounter troubles with saving some type of edits.

I'm using the version 19.0 on Snow Leopard, with 8Go of RAM, and the assembly has around 1037000 reads for a ~5Mb genome size. The ace file is loaded into Consed with all the phd files stored in a phd.ball for a faster start (1 min vs 15 hours). I can break contigs, join them, etc., I have no problem with saving newer versions of the ace files.

Recently, I've started editing the consensus sequence (by opening the 454 flowgrams and applying "Changing the consensus" for the selected bases). However when it comes to saving the edits, the process gets stuck, along with the error "exception thrown:" (without indication of what the error was, so not very helpful). I have the feeling that using the phd.ball to load phd data might cause the problem when saving edits, even though phd_dir is present and contains all individual phd files.

Has anyone encountered this problem before?

Thanks for helping.
Nouri

[flxlex]

Have you contacted the author(s) of consed (Phil Green)? I think the 454 compatibility is still very much under development...

[nbenzakour]

Thanks flxlex, I've contacted David Gordon in fact, I just thought I could ask the community in the meantime. That may also help others in the same situation.

[nbenzakour]

Hi all, thanks to David Gordon, here are the changes made to fix the problem:

1) The phd.ball was in edit_dir as it wouldn't open from phdball_dir. I modified the end of the ace file by adding a WA{} field specifying to consed to look for the phd.ball in phdball_dir and that did the trick.

This is how the WA field looks like. The time and date are not relevant by the way.

WA{
phdBall newbler 080416:144002
../phdball_dir/phd.ball
}

2) The phd_dir wasn't writeable (I can't believe I missed that).