taxonomy classification: genus Bacteroides belongs to family Lachnospiraceae?

[James Yang]

Hi there,

I downloaded the greengenes_tax_rdp_train_genus.txt.zip posted by Tony
Walters on Feb 23, 2012. I used it to analyse 16s rDNA 454
pyrosequencing data (length about 380 nts)  from intestinal digesta
samples. Although no phylum Bacteroidetes was not deteced by QIIME 1.4
, but  QIMME gave me about 10% genus Bacteroides. From my
understanding, genus Bacteroides should belong to phylum
Bacteroidetes. Then I checked the data base
greengenes_tax_rdp_train_genus.txt and found that line 309038
to 332586 say "Root;p__Firmicutes;c__Clostridia;o__Clostridiales;f__Lachnospiraceae;g__Bacteroides".
I don't know why genus Bacteroides is classfied to phylum Firmicutes.

Another question is that the output of taxonomy of QIIME is like
"Root;Root;p__Firmicutes;c__Clostridia;o__Clostridiales;f__Ruminococcaceae;g__Bacteroides".
Why two "Root" here?

Thanks,
James

[Tony Walters]

Hello James,

There still is some strangeness in the taxonomies due to older naming of microbial species with morphology/biochemical tests/colony morphology etc. that on occasion does not match up to phylogeny as dictated by DNA sequences.

Here is an example species of Bacteroides from the Firmicutes division:   http://genome.wustl.edu/genomes/view/bacteroides_capillosus 

I think there are some Clostridium genera that aren't in Firmicutes too.  It does create a headache for all involved.

As for the extra Root; showing up, I'm not totally sure-is that showing up in all of the assignments?  And you're running QIIME 1.4.0?

-Tony

[James Yang]

Hi Tony,

Thanks for your prompt reply.

Yes, I'm running QIIME 1.4.0 in Ubuntu Linux. The extra "Root;" shows
up in all of the assignments. Does this affect the final taxonomy
assignment?

For the first question, I searched Bergey's Manual of Systematic
Bacteriology and could not find the relationship between genus
Bacteroides and phylum Firmicutes.

thanks again,
James

On 5月9日, 下午4时55分, Tony Walters <william.a.walt...@gmail.com> wrote:
> Hello James,
>
> There still is some strangeness in the taxonomies due to older naming of
> microbial species with morphology/biochemical tests/colony morphology etc.
> that on occasion does not match up to phylogeny as dictated by DNA
> sequences.
>
> Here is an example species of Bacteroides from the Firmicutes division:http://genome.wustl.edu/genomes/view/bacteroides_capillosus
>
> I think there are some Clostridium genera that aren't in Firmicutes too.
> It does create a headache for all involved.
>
> As for the extra Root; showing up, I'm not totally sure-is that showing up
> in all of the assignments? And you're running QIIME 1.4.0?
>
> -Tony
>
>
>
> On Wed, May 9, 2012 at 2:28 PM, James Yang <xjyan...@gmail.com> wrote:
> > Hi there,
>
> > I downloaded the greengenes_tax_rdp_train_genus.txt.zip posted by Tony
> > Walters on Feb 23, 2012. I used it to analyse 16s rDNA 454
> > pyrosequencing data (length about 380 nts) from intestinal digesta
> > samples. Although no phylum Bacteroidetes was not deteced by QIIME 1.4
> > , but QIMME gave me about 10% genus Bacteroides. From my
> > understanding, genus Bacteroides should belong to phylum
> > Bacteroidetes. Then I checked the data base
> > greengenes_tax_rdp_train_genus.txt and found that line 309038
> > to 332586 say

> > "Root;p__Firmicutes;c__Clostridia;o__Clostridiales;f__Lachnospiraceae;g__Ba-cteroides".

> > I don't know why genus Bacteroides is classfied to phylum Firmicutes.
>
> > Another question is that the output of taxonomy of QIIME is like
>

> > "Root;Root;p__Firmicutes;c__Clostridia;o__Clostridiales;f__Ruminococcaceae;-g__Bacteroides".

[Tony Walters]

Hello James,

I don't think the extra root is affecting the final assignments themselves, as it is still going down to the genus level.  I could create some strangeness in the taxonomy plots from summarize_taxa_through_plots.py though (some graphs that are all just "root" assignments), but unless it was inconsistent in putting the extra root in front of the taxa string I don't think it will disrupt analyses.

You've probably already seen this, but we just released QIIME 1.5.0-it might be worthwhile to upgrade to see if this problem goes away (we'd like to know if there is something going with extra strings being appended in the current version).

-Tony