Question: How To Go Fishing In A Metagenomics Sample Using Raw Reads As Bait
0
gravatar for Lee Katz
6.8 years ago by
Lee Katz3.0k
Atlanta, GA
Lee Katz3.0k wrote:

Hi, I have a metagenomics sample. I also have a set of reads from a genome. How can I pull similar or identical reads from the metagenomics sample? My ultimate goal is to reconstruct the genome hidden in the metagenomics sample.

illumina metagenomics fastq • 2.1k views
ADD COMMENTlink modified 6.8 years ago by Pavel Senin1.9k • written 6.8 years ago by Lee Katz3.0k

I know that making an assembly and mapping would be one way to do it. This is a legitimate way but there will be some unassembled regions of the genome. What I'm asking is, is there a way to match reads to reads?

ADD REPLYlink written 6.8 years ago by Lee Katz3.0k

What exactly you would assemble, metagenome, or genome? If latter, then you will reduce the search space from a set of reads to contigs and singletons (= the index size) and speed-up the selection process. I am interested, why bwa wouldn't work? I use this solution in the current project and it works.

ADD REPLYlink written 6.8 years ago by Pavel Senin1.9k

After getting all the reads, I would assemble the genome hidden in the metagenome. The reason I don't want to map to an assembly is because some places in the assembly are not retained. Usually either due to repeats, misassemblies, or lack of coverage. I am willing to spend extra computer time on this in order to recover more or better reads.

ADD REPLYlink written 6.8 years ago by Lee Katz3.0k

Some good verbal suggestions to me have been either 1) find reads with identical kmers or 2) blast metagenomics reads vs genomics reads to recover any queries that match.

ADD REPLYlink written 6.8 years ago by Lee Katz3.0k
1
gravatar for Pavel Senin
6.8 years ago by
Pavel Senin1.9k
Los Alamos, NM
Pavel Senin1.9k wrote:

I use bwa to index the set of reads, then I align metagenomic sample to that reference - this will yield the "similar or identical reads".

ADD COMMENTlink written 6.8 years ago by Pavel Senin1.9k
0
gravatar for IV
6.8 years ago by
IV1.3k
USA
IV1.3k wrote:

How many are the reads from the genome that you already have?

If they are enough, you coud assemble them in contigs (or even to rough scaffolds of a genome) and then align against them.

If you have too few, the you could create a blast database or an aligner index and search against that.

Cheers,

IV

ADD COMMENTlink modified 6.8 years ago • written 6.8 years ago by IV1.3k

As I and Pavel already mentioned, you can create a bowtie, bwa or any other aligner index or blast database by using your available reads. You should at first collapse the reads and transform them into a fasta file. You can use this fasta to create your index or as the base of your blast database. You can use that index or that blast database to search against.

ADD REPLYlink modified 6.8 years ago • written 6.8 years ago by IV1.3k

I haven't used bowtie for quite a while but I think that the syntax could be something like:

bowtie-build myreadscollapsed.fa reads.index

ADD REPLYlink written 6.8 years ago by IV1.3k
Please log in to add an answer.

Help
Access

Use of this site constitutes acceptance of our User Agreement and Privacy Policy.
Powered by Biostar version 2.3.0
Traffic: 955 users visited in the last hour