Skip to main content

Cultivate your MSA to get better trees

Project description

Arbow: cultivate your multiple sequence aligment to get better trees


We named this tool arbow as that would be the phonetic pronounciation of the short, endearing, term for an arborist in Australia.

What it does

The goal of arbow is to automate and simplify the production of trees from multiple sequence alignments. The tool has been developed in the context of viral phylogenomics.

In the current version (0.1.0) it:

  1. Reads an alignment in multiFASTA format
  2. Calculates stats for each sequence in the alignment
  3. Calculates stats per column in the alignment
  4. Allows the user to set a threshold of tolerable missing data in a column, and removes all non-conforming columns from the alignment
  5. From the remaining columns, arbow finds all the constant columns, and calculates the frequency of A, C, G, and T sites (i.e., an alignment column with only A across all samples would count as 1 towards the overall frequency of As in the alignment).
  6. It then filters out all the variable columns, and outputs the variable alignment as a multiFASTA alignment.
  7. It runs IQTree with a few sensible presets

Currently, in step 4 above, columns that have a single observed nucleotide (e.g., C) but still have missing data that were not filtered out in step 3 are counted towards the overall frequency of that base in the alignment. In other words, if a user specifies a maximum number of 20 missing bases, and a column with 5 missing bases but with A in all other samples, that column will count towards the overall frequency of A in the alignment (i.e., majority consensus imputation). This assumptions is less risky the larger the number of samples in the alignment.

For step 5, missing data (i.e., - and N) are all codes as N.

Tests are underway to figure out how these assumptions might affect the output.


  1. Python >=3.6
  2. IQTree 1.6+ (not tested on IQTree 2 as it is not production ready yet)
  3. BioPython
  4. Pandas
  5. NumPy



brew install iqtree
pip<3> install arbow


conda install -c bioconda iqtree
pip<3> install arbow


  1. Generate a mulitple sequence alignment with your favourite aligner (e.g., MAFTT). Output a multiFASTA file.
  2. Run arbow <aln.fa>
  3. Open tree-YYYY-MM-DD_HHMMSS.treefile in your favourite tree viewer (e.g, FigTree)

Data stream

When running arbow, by default a stream is output to the console (stdout).

Data about the each sequence in the alignment is prefixed with [SEQ], and is followed by:

  1. Count of each base (A, C, G, T, and NN is any character other than ACGT)
  2. Percent missing data
  3. A status column that has 0, 1, 2, or 3 * depending on whether the percent missing data is <0.5, >=0.5 and <1.0, >=1.0 and <5.0, or >=5, respectively.

Data about each column in the alignment is prefixed with [ALN], and is followed by:

  1. Position in the alignment
  2. Count of each base (bases counted will depend on whether all IUPAC codes are allowed or not - see below in usage)

Command line


Usage: arbow [OPTIONS] ALN

  -i, --all-iupac               Print count of all IUPAC code for column

  -s, --no-stream               Stop streaming stats to console
  -mm, --max-missing INTEGER    Remove sites with 'mm' missing sites or more
                                [default: 20]

  -o, --out-var-aln TEXT        Filename for alignment of variable sites.
                                [default: aln-2020-04-07-150443.aln]

  -p, --prefix TEXT             Prefix to append to IQTree output files.
                                [default: tree-2020-04-07-150443]

  -t, --iqtree-threads INTEGER  Number of cores to run IQtree  [default: 4]
  -m, --iqtree-models TEXT      Substitution models to test.  [default:

  -f, --iqtree-freq TEXT        Base frequency models to test.  [default: F]
  -r, --iqtree-rates TEXT       Rate category models to test.  [default: G,R]
  -b, --iqtree-bb INTEGER       Maximum number of UltraFast Bootstrap
                                iterations to attempt.  [default: 1000]

  -a, --iqtree-alrt INTEGER     Number of replicates to perform SH-aLRT.
                                [default: 1000]

  -c, --iqtree-cmax INTEGER     Maximum number of rate categories to test.
                                [default: 5]

  --help                        Show this message and exit.

Get help

arbow --help

Get version

arbow --version

Project details

Download files

Download the file for your platform. If you're not sure which to choose, learn more about installing packages.

Files for arbow, version 0.1.3
Filename, size File type Python version Upload date Hashes
Filename, size arbow-0.1.3-py3-none-any.whl (6.9 kB) File type Wheel Python version py3 Upload date Hashes View
Filename, size arbow-0.1.3.tar.gz (7.0 kB) File type Source Python version None Upload date Hashes View

Supported by

AWS AWS Cloud computing Datadog Datadog Monitoring DigiCert DigiCert EV certificate Facebook / Instagram Facebook / Instagram PSF Sponsor Fastly Fastly CDN Google Google Object Storage and Download Analytics Microsoft Microsoft PSF Sponsor Pingdom Pingdom Monitoring Salesforce Salesforce PSF Sponsor Sentry Sentry Error logging StatusPage StatusPage Status page