Collapse Pango sublineages up to user defined parent lineages.
Project description
Pango-collapse
CLI to collapse Pango lineages up to user defined parent lineages for reporting.
Install
Install from pypi with pip.
pip install pango-collapse
Usage
pango-collapse
takes a CSV file of SARS-CoV-2 samples (input.csv
) with a column (default Lineage
) indicating the compressed pango lineage of the samples (e.g. output from pangoLEARN, nextclade, USHER, etc).
$ cat input.csv
Sample | Lineage |
---|---|
1 | BA.5.2.1 |
2 | BA.4.6 |
3 | BE.1 |
pango-collapse
will collapse lineages up to the first user defined parent lineage (specified in a text file with --collapse-file
). If the sample lineage has no parent lineage in the user defined collapse file the original compressed lineage will be returned.
The highest level you can collapse up to is either A
or B
. Additionally, there is a special Recombinant
lineage which can be used to collapse all recombinant lineages up to a single Recombinant
group.
By default (i.e. if no collapse file is specified) pango-collapse
uses the collapse file found here. This file is dependant on the version of pango-collapse
, use --latest
to load the latest version of the collapse file from github at run time.
Here is a simple example collapse file (collapse.txt) that will be used to collapse sublineages of BA.5
and BQ.1
.
$ cat collapse.txt
BA.5
BQ.1
pango-collapse
will produce an output file which is a copy of the input file plus Lineage_full
(the uncompressed lineage), Lineage_expanded
(the expanded lineage format) and Lineage_family
(the lineage collapsed up to) columns.
$ pango-collapse input.csv --collapse-file collapse.txt -o output.csv
Alternatively you can use the --parent
(-p
) flag to specify the parent lineages to collapse up to.
$ pango-collapse input.csv -p BA.5 -p BQ.1 -o output.csv
$ cat output.csv
Sample | Lineage | Lineage_full | Lineage_family | Lineage_expanded |
---|---|---|---|---|
1 | BA.5.2.1 | B.1.1.529.5.2.1 | BA.5 | B.1.1.529:BA.5.2.1 |
2 | BA.4.6 | B.1.1.529.4.6 | BA.4.6 | B.1.1.529:BA.4.6 |
3 | BE.1 | B.1.1.529.5.3.1.1 | BA.5 | B.1.1.529:BA.5.3.1:BE.1 |
Here we can see that samples 1 and 3 were collapsed to BA.5
. Sample 1 (BA.5.2.1
) and 3 (BE.1
) are both sublineages of BA.5
and BA.5
is returned in the Lineage_family
column. None of the parents of BA.4.6
(sample 2) are in the collapse file and so the orginal lineage is returned. None of the samples match BQ.1
(the other lineage in the collapse file) and so it is ignored.
Expanded lineage format
The Lineage_expanded
column contains the lineage in expanded format which provides complete lineage information and searchability.
B.1.1.529:BA.5.3.1:BE.1 == B.1.1.529.5.3.1.1 == BE.1
The delimiter (:
) separates each alias level in the full lineage. You can determine the parental lineages of a lineage in expanded format by reading from right to left. For example, from the expanded lineage B.1.1.529:BA.5.3.1:BE.1
we can see that BE.1
comes from BA.5.3.1
which inturn comes from B.1.1.529
.
Expanded lineages can be converted to full lineages by removing the delimiters and sub-lineage letters. Compressed lineages can be obtained by taking the final component of the expanded lineage.
$ echo "B.1.1.529:BA.5.3.1:BE.1" | sed -E 's/:[A-Za-z]+//g'
B.1.1.529.5.3.1.1 # full lineage
$ echo "B.1.1.529:BA.5.3.1:BE.1" | awk -F: '{print $NF}'
BE.1 # compressed lineage
Lineages to the right of a delimiter are equivalent (although the parental lineages are implicit).
B.1.1.529:BA.5.3.1:BE.1 == BA.5.3.1:BE.1 == BE.1
Lineages in expanded format are easily searched. Exact matches can be found by matching with the end of the expanded lineage e.g. using the regex $
anchor (:BE.1$
to exactly match BE.1). Sub-lineages can be found by simply checking if the expanded lineage contains the parental lineage of interest.
$ grep ":BA.5" output.csv # find all BA.5 sub-lineages
BA.5.2.1,B.1.1.529.5.2.1,BA.5,B.1.1.529:BA.5.2.1
BE.1,B.1.1.529.5.3.1.1,BE.1,B.1.1.529:BA.5.3.1:BE.1
Nextclade example
This example shows how to use some of the pango-collapse
features by collapsing the Pango Lineages in the output from Nextclade.
Produce a nextclade.tsv file from a nextclade
analysis (there is an example file in tests/data
).
We are only interested in the major sub-lineages of omicron i.e. BA.1-BA.5. We can therefor make a collapse file with the following:
$ cat collapse.txt
BA.1
BA.2
BA.3
BA.4
BA.5
Note: BA is an alias of B.1.1.529, however, as we have not included B.1.1.529 in our collapse file any samples designated B.1.1.529 will not be included.
Run the following command to collapse the omicron sub-lineages:
pango-collapse \
-c collapse.txt \
-o nextclade_collapsed_omicron.tsv \
-l Nextclade_pango \
--strict \
nextclade.tsv
The -l
(--lineage-column
) flag tells pango-collapse
to look for the compressed lineage in the Nextclade_pango
column in the nextclade.tsv file.
The --strict
tells pango-collapse
to use strict mode i.e. only report lineages in the collapse file. If the lineage cannot be collapsed then no value is returned in the collapse column.
We can visualise the results in pandas:
import pandas as pd
df = pd.read_csv("nextclade_output.tsv", sep="\t")
df.Lineage_family.fillna('Other', inplace=True)
df.Lineage_family.value_counts().plot(kind='bar')
Hosting collapse files
You can host your own collapse files and then use them with pango-collapse
to ensure that your reporting is always up to date. The simplest way to host a collapse file is on GitHub. You can then pass the url to pango-collapse
using the --collapse-file
(-c
) flag.
pango-collapse \
tests/data/input.csv \
-c https://raw.githubusercontent.com/path/to/your/collapse_file.txt \
-o collapsed.csv
This also has the advantage that non-technical users can update the collapse file without needing to interact with your reporting pipeline. For example, a genomic epidemiologist or government agency official could use the GitHub GUI to make a PR to your repository to add or remove lineages from the collapse file. Then the next time your pipeline runs it will pull in the updated collapse file.
Austrakka maintains Variant of Concern collapse files from several major public health agencies including the WHO, UKHSA, and the CDC. You can use these collapse files to report on the lineages of interest to these agencies.
# Collapse to the Variant of Concern lineages according to the CDC
pango-collapse tests/data/input.csv \
-c https://raw.githubusercontent.com/AusTrakka/VOC-watch/master/collapse_files/cdc.txt \
-o cdc_collapsed.csv \
--strict
--help
Usage: pango-collapse [OPTIONS] INPUT
Collapse Pango sublineages up to user defined parent lineages.
╭─ Arguments ─────────────────────────────────────────────────────────────────────╮
│ * input FILE Path to input CSV/TSV with Lineage column. │
│ [default: None] │
│ [required] │
╰─────────────────────────────────────────────────────────────────────────────────╯
╭─ Options ───────────────────────────────────────────────────────────────────────╮
│ --output -o FILE Path to output CSV/TSV with Lineage │
│ column. If not supplied will print to │
│ stdout. │
│ --collapse-file -c PATH Path or URL to collapse file with │
│ lineages (one per line) to collapse up │
│ to. Defaults to collapse file shipped │
│ with this version of pango-collapse. │
│ --parent -p TEXT Parental lineage to collapse up to. Can │
│ be used multiple times to collapse to │
│ multiple lineages. If --collapse-file │
│ is supplied parents will be appended to |
| the file. │
│ --lineage-column -l TEXT Column to extract from input file for │
│ lineage. │
│ [default: Lineage] │
│ --full-column -f TEXT Column to use for the uncompressed │
│ output. │
│ [default: Lineage_full] │
│ --collapse-column -k TEXT Column to use for the collapsed output. │
│ [default: Lineage_family] │
| --expand-column -e TEXT Column to use for the expanded output. |
| [default: Lineage_expanded] |
│ --alias-file -a PATH Path to Pango Alias file for │
│ pango_aliasor. Will download latest file │
│ if not supplied. │
│ --strict -s If a lineage is not in the collapse file │
│ return None instead of the compressed │
│ lineage. │
│ --latest -u Load the collapse from from a url │
│ (--url). │
│ --version -v Print the current version number and │
│ exit. │
│ --install-completion Install completion for the current │
│ shell. │
│ --show-completion Show completion for the current shell, │
│ to copy it or customize the │
│ installation. │
│ --help -h Show this message and exit. │
╰─────────────────────────────────────────────────────────────────────────────────╯
Project details
Release history Release notifications | RSS feed
Download files
Download the file for your platform. If you're not sure which to choose, learn more about installing packages.
Source Distribution
Built Distribution
Hashes for pango_collapse-0.8.2-py3-none-any.whl
Algorithm | Hash digest | |
---|---|---|
SHA256 | 4bd386e3c53d7c5d5071a59b048ce8efda7a51c498344c01544ae1b05b00262d |
|
MD5 | 8dcc8eb0fbfeb1bde9af110f2f8017b7 |
|
BLAKE2b-256 | b2fcbc926001cf1ee386702770021fc19b0864aa7620b50f5c86ea8403672315 |