Has anyone used the Genovo assembler?

[aroth8]

Today I found an article from Stanford about an assembler called Genovo, specially designed for assembling short reads (like the ones from the illumina sequencer) from metagenomes. Has anyone used this assembler? Is it good? I haven't read anything about it.
Thanks in advance

[rglover]

I've just read the paper, although the link for downloading the software keeps giving a 404 error... I'd love to give it a go on my metagenome datasets though to see how it compares to CLCbio too (they only compared Genovo, Euler, Velvet and Newbler).

[aroth8]

Thanks for your comment

rglover,

Thanks for your comment. Have you writen to the authors of Genovo? We are going to write them, maybe they could fix the problem of their webpage. Also I wanted to ask, this CLCbio assembler you mentioned is better than Velvet? Have you used Velved with your metagenomic data?
Thanks a lot, and hope to hear from you soon.

Alexandra

[rglover]

The website seems to be back up (http://cs.stanford.edu/genovo) - I'm just trying it out on a small metagenomic dataset (65000 seqs) to see how it compares on our data.
I've not tried velvet on our 454 datasets yet as there's never really been a need seeing as Newbler 2.5 and CLCbio have been doing a good job at assembling our (transcriptome) metagenomics samples (we're also assembling viral genomes from metagenomic samples from infected tissue). I'm intrigued by Genovo's results in the paper though - if it's producing much better assemblies, I don't mind it being a bit slower.

[Thomieh]

Hey rglover, I was wondering what your experience is with the genovo assembler?
I am running genovo myself on two different datasets to see how good the speed is, and how good it assembles with low coverage datasets.
I have tried Newbler and the clc de novo assembler which gave me very different results in the number of contigs.

[OHara]

Troubleshooting std::exception

Hi All,

There is understandably very little documentation for the Genovo assembler; Has anyone run into the following error? At the end of a run (100 iterations on a dataset of ~150,000 seqs) I see:

--------------------------------------------------------
setting costs to: indel=0 miss=0 unobserved=0 pad=0
setting costs to: indel=0 miss=0 unobserved=0 pad=0
setting costs to: indel=0 miss=0 unobserved=0 pad=0
init jvError Model
base insertion prob: 0.01 base deletion prob: 0.01 mismatch prob: 0.01
Loading State all_reads.fa.dump.best...state serializer: Reading data
state serializer: going over file...
jvStateSerializer::read sequence count parse failed
terminate called after throwing an instance of 'std::exception'
what(): std::exception
Abort
setting costs to: indel=0 miss=0 unobserved=0 pad=0
setting costs to: indel=0 miss=0 unobserved=0 pad=0
setting costs to: indel=0 miss=0 unobserved=0 pad=0
init jvError Model
base insertion prob: 0.01 base deletion prob: 0.01 mismatch prob: 0.01
Error: fasta file 120712-1.fa
terminate called after throwing an instance of 'std::exception'
what(): std::exception
Abort

Any help would be greatly appreciated

[doliver]

STD::Exception Error with Genovo Assembler

I've run into the same error as you have by using the DEMO.sh to set up my run. If you have solved the issue, I'd be curious how you went about resolving the problem.

[OHara]

Actually doliver, in my case it was quite a simple mistake; I had not modified all of the necessary parts of the DEMO.sh-based script I was using for the run. Specifically, my input fasta was not named all_reads.fa, and so I needed to make that change to the "finalize" input. Fortunately this is after "assemble" has already completed and so I don't believe the run has to be repeated. I just commented-out the "assemble" from the script and re-ran with the "finalize" change.

Hope your fix is as simple!

[doliver]

Thanks!

Thanks for replying. I don't expect that is my error since I'd modified the script with new input and output file names and though the script ran for a long time, no output file was generated but rather I encountered an error. I even tried running just the assemble program on its own and it failed as well. How was your assembly? I was about to give up on Genovo but if the assembly is better, I might give it another shot.

[OHara]

Once I fixed my mistake, the output was great. The contigs match well to anticipated genes (and have neighbouring genes with related functions.) I don't know of a good database or scaffold to compare what I'm looking at (rat intestinal microbes) but from what I can glimpse so far, it's great. My dataset was also not originally obtained with genome assembly in mind, so my biggest contigs are only 20 kb, but I'm not sure if there's anything I can do about that.

[rsinha]

How to use Genovo for paired-end data?

Hi

I could not find a way to use paired end data while running Genovo assembler, as it takes only single reads file?
I will be glad to have answer.

Thanks