News and Announcements » – Filter sequence alignment by removing highly variable regions


This script should be applied to generate a useful tree when aligning against a template alignment (e.g., with PyNAST). This script will remove positions which are gaps in every sequence (common for PyNAST, as typical sequences cover only 200-400 bases, and they are being aligned against the full 16S gene). Additionally, the user can supply a lanemask file, that defines which positions should included when building the tree, and which should be ignored. Typically, this will differentiate between non-conserved positions, which are uninformative for tree building, and conserved positions which are informative for tree building. FILTERING ALIGNMENTS WHICH WERE BUILT WITH PYNAST AGAINST THE GREENGENES CORE SET ALIGNMENT SHOULD BE CONSIDERED AN ESSENTIAL STEP.

Usage: [options]

Input Arguments:



-i, --input_fasta_file
The input fasta file containing the alignment


-o, --output_dir
The output directory [default: .]
-m, --lane_mask_fp
Path to lane mask file [default: 16S alignment lane mask (Lane, D.J. 1991)]
-s, --suppress_lane_mask_filter
Suppress lane mask filtering [default: False]
-g, --allowed_gap_frac
Gap filter threshold, filters positions which are gaps in > allowed_gap_frac of the sequences [default: 0.999999]
-r, --remove_outliers
Remove seqs very dissimilar to the alignment consensus (see –threshold). [default: False]
-t, --threshold
With -r, remove seqs whose dissimilarity to the consensus sequence is approximately > x standard deviations above the mean of the sequences [default: 3.0]
-e, --entropy_threshold
Percent threshold for removing base positions with the highest entropy, expressed as a fraction between 0 and 1. For example, if 0.10 were specified, the top 10% most entropic base positions would be filtered. If this value is used, any lane mask supplied will be ignored. Entropy filtering occurs after gap filtering. [default: None]


The output of consists of a single FASTA file, which ends with “pfiltered.fasta”, where the “p” stands for positional filtering of the columns.

Example 1:

As a simple example of this script, the user can use the following command, which consists of an input FASTA file (i.e. resulting file from and the output directory “filtered_alignment/”: -i seqs_rep_set_aligned.fasta -o filtered_alignment/

Example 2:

Apply the same filtering as above, but additionally remove sequences whose distance from the majority consensus sequence is more than 3 (can be changed by passing –threshold) standard deviations above the mean: -i seqs_rep_set_aligned.fasta -o filtered_alignment/ --remove_outliers

Example 3:

Alternatively, if the user would like to use a different gap fraction threshold (“-g”), they can use the following command: -i seqs_rep_set_aligned.fasta -o filtered_alignment/ -g 0.95