I've been doing some research into simulators for simulating metagenomes and I cannot find a good consensus on what to use. The two big ones I've seen are MetaSim and BEAR. Does anyone have experience with these or others or have any advice for a good meta-genome simulator?
•
written
by
dlawre14 •
BBMap's simulator, randomreads.sh, has a metagenome mode; just add the flag "metagenome". E.g.
cat bug1.fa,bug2.fa,bug3.fa > bugs.fa
randomreads.sh ref=bugs.fa out=reads.fq reads=10m len=150 paired metagenome
It also has another tool, "mutate.sh", to create strains from a reference with slight differences. This can be useful when simulating metagenomes.
I hadn't thought of BBMap... I swear that thing does everything now. Thank you!
[As of BBmap v. 36.59] randomreads.sh gains the ability to simulate metagenomes.
coverage=X will automatically set "reads" to a level that will give X average coverage (decimal point is allowed).
metagenome will assign each scaffold a random exponential variable, which decides the probability that a read be generated from that scaffold. So, if you concatenate together 20 bacterial genomes, you can run randomreads and get a metagenomic-like distribution. It could also be used for RNA-seq when using a transcriptome reference.
The coverage is decided on a per-reference-sequence level, so if a bacterial assembly has more than one contig, you may want to glue them together first with fuse.sh before concatenating them with the other references.
OMG, BBmap can do anything!
You would do something like this:
cat bacteria1.fa bacteria2.fa (and so forth) > all.fa
randomreads.sh ref=all.fa out=reads.fq len=150 paired reads=10000000 metagenome
Then it will generate reads with different coverage for each sequence in the reference.
Is there anyway to run this so that it's species/isolate aware? The model would be more representative of real communities if it assigned probabilities to species/isolates.
Log in to answer this question.
