classify.seqs - not sufficient taxonomy level

mothur bugs
1

Hi, Im trying tu run classify.seqs command to classify my 18S v9 domain data. I downloaded Silva nr alignment v 123 and corresponding taxonomy file. Than I aligned my data to the silva alignment to get start and end position. Then I used pcr.seq command to get only V9 domain sites.
when I run classify.seqs command, the Mothur do classification to the 6th level (regardless what the level is). How to get information for lover taxonomic levels?

Well, its good to know, that the sequence A is an ascomycota, but it would be much better to know which genus of ascomycota… :slight_smile:

2

What does the full output for the sequence assigned to ascomycota look like in your taxonomy file?

Also can you paste the actual command you ran.

3

Well, I tried classify.seqs with SILVA nr 123 template (both 1-full and 2-trimmed to amplicon size)
same result->just 6 levels

I used a toy sequence : genbank record is AF006309.1 (trimmed to my amplicon size)
my toy amplicon looks like this:

gi|5869993|gb|AF006309.1||r
GCTACTACCGATTGAATGGCTCAGTGAGGCCTTCGGACTGGTCCAAAGAGGTCGGCAACGGCTTTTTCAGGTGCTGGAAAGTTGGTCAAACTTGGTCATTTAGAGGAAGTAAAAGTCGTAACAAGGTTTCCGTAGGTGAACTGCGGAAA

result looks like this:
gi|5869993|gb|AF006309.1|
Eukaryota(94);Opisthokonta(64);Nucletmycea(52);Fungi(48);Dikarya(42);Ascomycota(42);

The result from my real amplicon data looks in the same way->just 6 taxonomic levels.

the command was:
mothur > classify.seqs(fasta=toysample.fasta, taxonomy=silva.nr_v123.tax, template=silva.nr_v123.ng.fasta)

4

A little update

When I used older template, it looks better.
I downloaded Eukaryotic references of the 102 release from http://www.mothur.org/wiki/Silva_reference_files

my command was:
mothur > classify.seqs(fasta=toysample.fasta, taxonomy=silva.eukarya.silva.tax, template=silva.eukarya.fasta)


the result was:

gi|5869993|gb|AF006309.1||r Eukaryota(100);Fungi(100);Dikarya(100);Ascomycota(100);Pezizomycotina(100);Pezizomycetes(99);Pezizales(99);Pezizaceae(99);Peziza(99);unclassified;unclassified;unclassified;unclassified;unclassified;unclassified;unclassified;unclassified;unclassified;unclassified;

It looks much better, but there is a lot of “unclassified” after the Genus name. It looks weird.

5

So if you open your silva.nr_v123.tax file you will see that it only contains six levels of annotations which is why you can’t get anthing beyond Ascomycota. The 102 version of the mothur compatiable SILVA alignment was likely constructed differently giving you the different results you see.

I personally haven’t worked with 18S data in mothur so I don’t know what the best option is, but you may benifit from using a different reference database for classifying, or creating a custom reference database that meets your needs better. To get an idea of how to go about this you can look at the README’s for the mothur compatible SILVA/Greengenes releases which document their creation.

Cheers
Richard