SOAPdenovo2 assembly problem at contig step

[Hari]

Hi,

For the assembly using sparse pregraph step we get the following when the edge construction step commences:

Start to construct edges.

INFO: two loop edge sequence are reversed complemental!

GAATATAAAGCTTTATAGGTTACCTCATCCTATAAAGCTTTATATTC

GAATATAAAGCTTTATAGGATGAGGTAACCTATAAAGCTTTATATTC

INFO: two loop edge sequence are reversed complemental!

TTGGGTCAGTTCTGATGAGACCCCCGTTCTCATCAGAACTGACCCAA

TTGGGTCAGTTCTGATGAGAACGGGGGTCTCATCAGAACTGACCCAA

INFO: two loop edge sequence are reversed complemental!

AGTACAGACTCCGGAGCTAGCTGGCAGCTAGCTCCGGAGTCTGTACT

AGTACAGACTCCGGAGCTAGCTGCCAGCTAGCTCCGGAGTCTGTACT

INFO: two loop edge sequence are reversed complemental!

CCATGTGTAAATTCTCTATAGTAGGTATATAGAGAATTTACACATGG

CCATGTGTAAATTCTCTATATACCTACTATAGAGAATTTACACATGG

INFO: two loop edge sequence are reversed complemental!

GTCCTTCACCAGATTTATGGCTGAAGGACATAAATCTGGTGAAGGAC

GTCCTTCACCAGATTTATGTCCTTCAGCCATAAATCTGGTGAAGGAC

INFO: two loop edge sequence are reversed complemental!

TATTTGGAATATAATCAACAGGATTCCTGTTGATTATATTCCAAATA

TATTTGGAATATAATCAACAGGAATCCTGTTGATTATATTCCAAATA

INFO: two loop edge sequence are reversed complemental!

TCACCTGGGGTTCTAACTCGTCTTCCGAGTTAGAACCCCAGGTGA

TCACCTGGGGTTCTAACTCGGAAGACGAGTTAGAACCCCAGGTGA

409641430 edges and 67081046 vertexes constructed.

Time spent on constructing edges: 3929s.

And when we attempt to contig the graph the following is the output with no contig file being formed. Could you kindly tell us what is the problem that we're facing?

We have preprocessed the reads using prinseq at Q30 and removed the duplicates.

Version 2.04: released on July 13th, 2012

Compile Jun 13 2013     16:55:53

********************

Contig

********************

Parameters: contig -g /share/apps/XXXXX/assemblies/soap/kmer17/soap -R

There are 67081046 kmer(s) in vertex file.

There are 409641430 edge(s) in edge file.

Kmers sorted.

409641430 edge(s) input.

521350278 pre-arcs loaded.

102419649 none-palindrome edge(s) swapped, 0 palindrome edge(s) processed.

ATTAACTATCCACAGGG

ATTAACTATCCACAGGG

TCTTCAAGACCCT

TCTTCAAGACCCT

cmp_seq:        err

Thanks,

Hari

[Snehal Karpe]

Hi,

I am encountering the same "cmp_seq:        err" error, when I use 'm' option for multiple kmers.
For single kmer 63, it is working fine.

I am giving following command: soapdenovo/pregraph_sparse_127mer.v1.0.3 -s ABC.config -K 15 -m 83 -z 3000000000 -R -p 16 -o XY 1>pregraph.log 2>pregraph.err
but in pregraph.err it shows it as follows: pregraph_sparse -s ABC.config -K 15 -g 15 -z 3000000000 -d 1 -e 1 -R  -r 0 -p 16 -o XY
(So I am not sure whether it is using 'm' option or not.)
Here I get error like ("INFO: two loop edge sequence are reversed complemental!.... as mentioned in the previous mail by Hari").
Does having reads of different length (in my case, 85 or 86 or 87) affect the procedure?

Then I run: soapdenovo/SOAPdenovo-127mer contig -s ABC.config -g XY -R -p 16 1>contig.log 2>contig.err
and in contig.err file it shows as:contig -s ABC.config -g XY -R -p 16

I am not able to figure out where am I going wrong.
Please help.

Thank you and regards,
Snehal

[lamz1...@163.com]

Hi!

I had the same problem, did you solve it? With subset of reads, no error was reported, so I thought it may out of memory. After tracing the memory usage, I exclude this possibility. Any suggestion would be grateful!

Following is my log:

41306549 none-palindrome edge(s) swapped, 0 palindrome edge(s) processed. AGGAATGGAAGGTCATAATTATA AGGAATGGAAGGTCATAATTATA TTGAAAATATAATTA TTGAAAATATAATTA cmp_seq: err

在 2014年1月13日星期一 UTC+8下午1:58:37，Hari写道：

[Snehal Karpe]

I encountered the problem for multiple k-mer option. It was not resolved.

[phillip kyriakakis]

Me too, still not resolved. Anyone figure it out?