CLC-Reads do not map back to contigs

[vanillasky]

I have been using CLC genomics to try and assemble paired end reads from an illumina run (this is an environmental metagenome sample). The assembly is set up so that at the end, the reads are mapped back to the contigs to see how many matched and how many did not match the contigs. I see from our output that a large number of reads do not match the contigs. Is this indicative of misassembly or what could this indicate? Any input would be most appreciated.

p.s. the trimming was done with an adapter input file to remove any of these present in the sequences. We see only a small percentage having undergone adapter removal following the trimming process. The majority of sequences were just trimmed based on quality score.

[GenoMax]

This would be a good candidate to contact CLC Tech support about. They should be able to remote in and see what is going on. I doubt we can give you any good advice here (as far as CLC goes).

[vanillasky]

Hi GenoMax,

I actually obtain similar results using other assembly programs such as velvet for example. Why would a high number of reads in general not match up with a set of contigs post assembly?

[GenoMax]

Was that with metavelvet (http://metavelvet.dna.bio.keio.ac.jp/) or regular velvet? Since this is a metagenomic sample you should use an assembler that is designed for that purpose. Ray-meta, MetAMOS look like other options.

[yzzhang]

for metagenome assembly, is it normal that a large percentage of reads can not be assembled to contigs, right?