Maximum number of contigs limit

[Ariel Paulson]

Hi,

I need to visualize the Python genome, which is at 1.3M contigs, but
the current limit is 1M. Can the limit on the dev version be raised?

There's not an easy way for me to winnow down the contig count since
gene models are found across the full range of sizes, and who knows
which contigs will receive future blast hits. I'd rather not create a
bunch of fake scaffolds just to reduce the sequence count, unless IGV
would really drag with that many contigs anyway.

Thanks,
Ariel

[Jacob Silterra]

I've just increased the MAX_CONTIG value to 1.5 million in the earl
access version. Let me know if it performs alright.

In the future, I'm not sure how high we want to push this parameter in
the official distribution. If you don't mind building IGV yourself
it's pretty easy to change; it's just a constant in GenomeImporter.
More detailed build instructions are in readme_src.txt, but simply
executing "ant" (after installing ant) from the igv source directory
should give you a usable jar.

-Jacob

--
Jacob Silterra
Software Engineer
Broad Institute

[Ariel Paulson]

Thanks Jacob!

I had posted here because I tried changing the source but it won't
recompile properly, so I figured someone here would implement the
change before I did...

The build.xml specifies a lot of things that don't seem to come from
the source distrib, at least not from the unpacked igv.jar. There is
no readme_src.txt that I can find. Where do you get it?

Thanks,
Ariel

[Jacob Silterra]

What kind of error message did you get when building?

readme_src is not in the binary distribution, only in the source. It
should be "docs" directory:

igv_root/
docs/
readme_src.txt
readme.txt
.....
src/
scripts/
.....

There isn't too much there right now beyond what I already said, but
I'll be maintaining a list of problems people encounter while building
and how to solve them in that document.

-Jacob

--

[Ariel Paulson]

I downloaded the source from 'IGVDistribution_2.0.34.zip' and this
seems to have everything.

Previously I had downloaded 'IGVDistribution_dev.zip' and I guess
something went wrong with the unpack, because I only got the 'src'
dir, and not all the others -- but re-unpacking it shows all the stuff
ant was missing.

I should be good from here.

Thanks,
Ariel

[Jacob Silterra]

I've been adjusting the build process recently to streamline it, I
guess I missed a few things.

You can always get the latest code at the subversion repo (link on the
IGV download page) as well.

--

[Jim Robinson]

Hi,

Back to the original request, to increase the maximum # of contigs past a million.   I just checked in an important update for the "import genome" option,  in the past it would create a separate file for each sequence in the fasta, which would obviously be a problem in cases like this and was annoying in all cases.   It no longer does this.   The new behavior is to first look for an index (.fai) file, and if one is not found it will create it.     This is in the "early access" version and will be released with 2.1 in the next couple of weeks.

Jim