Question about 'masked' alignment- how do I acquire details of alignment?

I am using Qiime2 2019.4; I can update if necessary. I have put some 16S sequencing data into alignment via mafft alignment algorithm prior to diversity analyses. The data includes V4 amplicons (515-806) and V3V4 amplicons (341-805) from different Illumina runs. I understand the limitations and ramifications of this approach, but nevertheless, I am curious about the ‘masked’ alignment that has been generated. The masking, I understand, is performed to remove ambiguous portions of sequences. I am guessing that this means that non-homologous portions of the sequences are removed, but I am really not at all sure what is actually going on. To help answer my question of what is going on, I would like to inspect the masked alignment (qza) and the unmasked alignment to see how they compare to the rep-seq files. I would like to see what data remains following masking and determine (or estimate) the average sequence length following masking. I have found that the ‘export’ function does not work on alignment files, there does not seem to be a way to convert alignment to visualizations, and I do not know of an external program that can receive alignments (in qza format) and visualize them. So, is there a way to determine what the masking is doing? Any thoughts or insights on these curiosities would be much appreciated.

Hello Nick,

Instead of ‘exporting’ your data, try ‘extracting’ it. Like this:

mkdir extracted-data
qiime tools extract \
  --input-path your-file.qza \
  --output-path extracted-data

The directory should have the MSA in a fasta format natively produced by MAFFT.

Masking is just a way to tell programs to ignore parts of the sequence, but which parts are being ignored are up to you. So you could mask out repetitive regions (ATATATATATATATATATA) or you could mask out conserved regions. What kind of masking did you perform here?

Colin

P.S.

:+1:

Great, thank you!

I had just followed the default ‘qiime alignment mafft’ to make the alignment with no particular intended masking. I’ll look over the alignments and then be able to describe what they look like better.

Thank you, this really helps me quite a bit.

Nick

This topic was automatically closed 31 days after the last reply. New replies are no longer allowed.