align.seqs using silva.nr_v123

Theory behind mothur
1

Hi,
I want to align my 16S sequences using the latest silva reference alignment (v123). For the align.seqs command, is reference=silva.nr_v123.align? I hesitate because the the Silva reference files page on the wiki (http://www.mothur.org/wiki/Silva_reference_files) says that this reference could be customized for alignments. Do I need to customize the file, or can I use it directly? In the past, I have used silva.bacteria.fasta from 2012 as the reference file, but now I would be using a .align file as the reference file.

After aligning to silva.nr_v123.align, my summary.seqs() output:

Start End NBases Ambigs Polymer NumSeqs
Minimum: -1 -1 0 0 1 1
2.5%-tile: 21597 27659 201 0 4 4364
25%-tile: 21787 27659 203 0 4 43631
Median: 21929 27659 209 0 4 87262
75%-tile: 21934 27659 220 0 5 130893
97.5%-tile: 21969 28450 234 0 6 170160
Maximum: 43116 43116 355 0 8 174523
Mean: 21874.7 27805.3 211.598 0 4.28712

of unique seqs: 33904

total # of seqs: 174523

Besides the 1 sequence that starts and ends at -1, the other positioning looks okay to me. Do you agree?

2

Hi Nadine,

You can use either the nr or the seed version of the reference alignment. Your alignment looks pretty good. I would probably set start=21969 and end=27659 in screen.seqs.

Pat

3

Maybe it is not the best way to ask.

I am trying to use the RDP trainset14 to align my sequences using align.seqs and it aborts telling me that the template is not aligned.

now I’am tryin to fix this using clustalW to align the reference file and try again. If there is another solution for this problem I accept suggestions.
Best regards,
Felipe

4

Hi,

Please post your question as a new thread. The RDP reference files are not aligned (the alignments they provide are pretty bad) and we strongly encourage people to use the silva reference files.

Pat

5

Thank you Pat
I’ll open another thread.