Closed krithikaarumugam closed 6 years ago
I don't think it makes sense to specify -nanopore-raw
for the corrected reads, -nanopore-corrected
would run what you need (trim/assembly only). Given that it works on raw input reads and we've made some significant changes in this code post the 1.6 release, I'd rather hold off on troubleshooting the crash and recommend you stick with -nanopore-raw
for uncorrected and -nanopore-corrected
for nanocorr reads.
As for the second question, the genome size isn't really well defined for a metagenome but it is used by Canu to select the longest reads for correction. In your case, you'd be losing about 50% of your coverage. However, if the metagenome is not very complex it may be reconstructing the most abundant organism fine without the extra data. See the FAQ: http://canu.readthedocs.io/en/latest/faq.html#what-parameters-can-i-tweak for our current suggested parameters for metagenomes to get all the data corrected, regardless of genome size. The genome will then just be used for NG stats computations.
Thanks for the quick reply. Was just trying to take advantage of the option -nanopore-raw incase any reads were missed during correction.
If the input is the data from nanocorr canu won't ever see the uncorrected reads and the consensus will be based completely on the nanocorr reads so I don't think any of them are going to be improved much and it won't find reads that weren't corrected since those wouldn't have been output by nanocorr. You can compare the uncorrected raw data assembled with -nanopore-raw
(without nanocorr) to the -nanopore-corrected
(with nanocorr) and see which gives a better assembly and use that one.
Inactive, reads corrected by other tools (nanocorr, lordec, etc) should typically be input as nanopore-corrected
or pacbio-corrected
not raw.
Hi,
I've been trying use can-1.6 on metagenomic nanopore reads corrected using
nanocorr
. I seem to get the following error (log file pasted below). Am not sure if am missing the obvious.Below is the command I used on Linux.
The following is the log file,
Also when I try to use only the -trim and -assemble options for corrected nanopore reads, the job seems to complete but I get the same assembly statistics (N50, mean etc) when I vary the genome size. am I missing the obvious here ?
P.S. - the job did finish successfully if I use the same command on uncorrected nanopore reads (without using nanocorr).
Thanks In advance for any help.