Re: [wgs-assembler-users] small gaps of fixed length

[Sajeet H. <sa...@gm...>]

Hello Brian,

Since all the CTP mea is more than -20, can I assume that no overlaps were
skipped where contigs overlapped more than 20 bases (ie: all overlaps > 20
were successfully merged into scaffolds with no N's).

How does the scaffolder behave when the overlap between contgs is 5bp or
less?

For a small fungal genome with relatively little repeats and no allelic
variations (is haploid), are there any parameters the can reduce the number
of false gaps. Will increasing cgwErrorRate help?

Thank you,

Sajeet


-----Original Message-----
From: Walenz, Brian [mailto:bw...@jc...] 
Sent: May-16-12 7:35 PM
To: Sajeet Haridas
Subject: RE: small gaps of fixed length

IIRC, that's the marker for "contigs should overlap, but no overlap found".
Possibilities here: the overlap is shorter than we can detect, or there is
crud on the end of one contig, or the error rate is too high.

Are you on the mailing list?  This would have been a nice discussion there.


________________________________________
From: Sajeet Haridas 
Sent: Wednesday, May 16, 2012 8:22 PM
To: Walenz, Brian
Subject: RE: small gaps of fixed length

Thank you Brian.
I also notice that the minimum CTP mea is -20. Is this value also capped?

Sajeet



From: Walenz, Brian 
Sent: May-16-12 1:12 PM
To: Sajeet Haridas
Subject: Re: small gaps of fixed length

Yes - that's the lower limit on a gap between contigs.  Either mate pairs
indicate the contigs should overlap but no overlap could be found, or there
really is a small positive gap.  Only the asm file will distinguish the two.

Look under 'mea' here:
http://sourceforge.net/apps/mediawiki/wgs-assembler/index.php?title=ASM_File
s#SCF_CTP

b


On 5/16/12 3:29 AM, "Sajeet Haridas"  wrote:
Hello Brian,

My fungal genome assemblies (30-35MBp) always seem to have ~2500 small gaps
always represented by 20 N's - using bog, bogart and various other
parameters.
Is the assembler trying to tell me something?

Thank you,

Sajeet