SOAP Scaffolding Inconsistencies

[kaboroevich]

Hello all,

Sorry if this is a repost. I was sure I posted this last week but I cannot find the original using google groups search.

I am getting some conflicting results for the SOAPdenovo scaffolding process. I found that if I use the  ouput of the prepare (http://soap.genomics.org.cn/down/prepare.tgz) tool, I get different, arguably better, results.

After error correction (http://soap.genomics.org.cn/down/correction.tar.gz), I am left with 242,857,233 reads in 113 libraries. Running pregraph and contig with a kmer of 49, I generate “original.contig”. Following the normal steps, I ran map and scaff resulting in “original.scafSeq”.  I also copied “original.contig” to a new directory and ran “prepare –g rescaf –K 49 –c original.contig”. Running map and scaf on rescaf.contig results in the file rescaf.scafSeq. The following is a summary of the assemblies.

File              Contigs  Total_length Longest   N50

original.contig   2459270     144845786    5254    54

rescaf.contig     2459270     144845786    5254    54

original.scafSeq    15289       4964403    5254   532

rescaf.scafSeq       9875       5416552   36403  2579

Does anyone know the explanation for this difference? The only difference between “original.contig” and “rescaf.contig” is the lack of “cvg_X.X_tip_X” in the latter. If any information on the accompanying files is needed, I can provide that as well.

Regards, Keith