[mira_talk] Re: Discrepancy in numbers
- From: Torben Nielsen <torben@xxxxxxxxxx>
- To: mira_talk@xxxxxxxxxxxxx
- Date: Fri, 6 Dec 2013 15:19:54 -1000
I’m afraid I must leave it to higher authority to answer your prayers. It’s a
marine metagenomic assembly. It’s a 25m depth one. I did over 20 different
assemblies and I noticed significant variation (factor of two) in the numbers
as reported in the assembly statistics file and as given by the contig
statistics file. You said that the latter was authoritative so that’s what I
have been using. I was going to write a script to extract N50 and N90 from
that, but I haven’t gotten to it yet.
I just finished a pooled run of over 20 datasets all from the same location
over time. It took almost two weeks and 256 GB of memory, but it came out
nicely. But it just finished yesterday so I haven’t really checked the numbers.
Thanks, Torben
On Dec 6, 2013, at 11:09, Bastien Chevreux <bach@xxxxxxxxxxxx> wrote:
> On 24 Nov 2013, at 1:30 , Torben Nielsen <torben@xxxxxxxxxx> wrote:
>> […]
>> The assembly statistics file is a nice summary to have. But couldn’t it be
>> generated by a script from the contig statistics? Mostly anyway…..
>
> Ummm, what you report there looks like some dumb bug somewhere. Pray tell,
> was that an EST/RNASeq assembly?
>
> B.
>
>
Other related posts:
