Amplicon analysis is not much different from standard phylogenetic placement, see here for instructions on that. However, in amplicon analysis query sequence are generated with PCR and are thus expected to match a specific target region in the reference alignment. Typically only the region that the query sequences match and only the reference sequences closely-related to the query sequences are of interest. To facilitate amplicon analysis, PAGAN contains options to trim and prune the resulting alignment.
The extra options for amplicon analysis are bets explained with an example. The data used in the examples below can be found here.
The following files are included:
- pagan_placement.cfg : config file that runs a basic analysis;
- query.fas : amplicon sequences to place in the alignment;
- reference.fas : reference alignment (these from Silva);
- reference.tree : reference phylogeny.
The data can be analysed using the following command:
The same analysis could be performed with the command:
pagan --ref-seqfile reference.fas --ref-treefile reference.tree \ --queryfile query.fas --outfile pagan_placement --xml \ --one-placement-only --trim-extended-alignment \ --prune-extended-alignment --prune-keep-number 0 \ --prune-keep-closest
This produces several output files:
- pagan_placement.??? contain the full data;
- pagan_placement.trimmed.??? contain the target region for all the sequences;
- pagan_placement.pruned.??? contain
the target region for the query sequences and for N reference sequences (here N=0);
- pagan_placement.pruned_closest.??? contain the target region for
the query sequences and the very closest reference sequences.
Files ending with .tre are the alignment guide trees for each output; those ending
with .fas and .xml are the alignments in FASTA and HSAML format.
If the amplicon sequences are very closely related to the reference sequences, one can
add option --terminal-nodes and thus enforce their placement next to the tips.
Options related to trimming and pruning are:
- prune-extended-alignment : prune output;
- prune-keep-number arg : prune output and keep N most distantly related sequences (if = 0, keep only query sequences);
- prune-keep-threshold arg : prune output and remove sequences with distance below threshold;
- trim-extended-alignment : trim reference alignment outside query target region;
- trim-keep-sites arg : trim distance around query sequences.
prune-keep-closest : prune output and keep only closest reference sequences;
The .xml output files can be visualised with Wasabi.