https://github.com/gamcil/fungiphy
Fungal marker-based phylogenetics toolkit
https://github.com/gamcil/fungiphy
Last synced: 11 months ago
JSON representation
Fungal marker-based phylogenetics toolkit
- Host: GitHub
- URL: https://github.com/gamcil/fungiphy
- Owner: gamcil
- License: mit
- Created: 2020-03-05T06:49:24.000Z (over 6 years ago)
- Default Branch: master
- Last Pushed: 2025-04-24T17:03:07.000Z (about 1 year ago)
- Last Synced: 2025-05-08T23:49:27.598Z (about 1 year ago)
- Language: Python
- Size: 2.44 MB
- Stars: 1
- Watchers: 1
- Forks: 0
- Open Issues: 1
-
Metadata Files:
- Readme: README.md
- License: LICENSE
Awesome Lists containing this project
README
# fungphy
Fungal marker-based phylogenetics toolkit
## Usage
Clone, set up virtualenv, install fungphy
```sh
git clone https://github.com/gamcil/fungphy.git
cd fungphy
python3 -m virtualenv venv
source venv/bin/activate
pip3 install -e .
```
### Setup
Initialise database
```python3
>>> from fungphy.database import db
>>> db.create_all()
```
Scrape aspergilluspenicillium.org for Aspergillus, Penicillium & Talaromyces markers
```python3
>>> from fungphy import scraper
>>> good, bad = scraper.scrape()
>>> with open("table.csv", "w") as fp:
... for sp in good:
... columns = ",".join(sp)
... fp.write(columns)
```
Import .csv
```python3
# Genus|Species|Reference|MycoBank ID|Type|Ex-types|Subgenus|Section|*markers
>>> from fungphy import importer
>>> with open("table.csv") as fp:
... importer.parse_csv(fp)
```
### Express usage
```python3
>>> from fungphy import plot
>>> table, msa, tree = plot._run(
... genera=["Aspergillus"],
... species=["avenaceus", "alliaceus", "albertensis", "tamarii", "caelatus",
... "oryzae", "flavus", "nomius", "bombycis", "coremiiformis", "togoensis"],
... markers=["ITS", "BenA", "CaM", "RPB2"],
... trim_msa=True,
... run_fasttree=True, # Generate tree using FastTree
... show_tree=True,
... outgroup="avenaceus", # Root this tree on A. avenaceus
... bold=["nomius", "bombycis"], # Bolden leaf labels for A. nomius and A. bombycis
... types=["albertensis", "tamarii"] # Use holotype strain name
... )
```
Generates:
Maximum likelihood tree inferred from combined ITS, BenA, CaM and RPB2 sequences of taxa
within subg. *Circumdati* sect. *Flavi* using FastTree.
### Alignments & phylogeny
Currently, procedures are in the `plot` module, so import that
```python3
from fungphy import plot
```
Query DB for species
```python3
>>> species = plot.get_species(
... genera=["Aspergillus"],
... species=["avenaceus", "alliaceus", "albertensis", "tamarii", "caelatus",
... "oryzae", "flavus", "nomius", "bombycis", "coremiiformis", "togoensis"],
... )
>>> species
[, , , , , , , , , , ]
```
Note that the returned objects are instances of the `Strain` class. The model hierarchy
is as follows:
```
Subgenus
Section
Genus
Species
Strain
Marker
```
Since the species descriptions on aspergilluspenicillium.org describe just the type
and ex-types formally attached to the species name, there will only be one `Strain` per
`Species`. If more `Strain` objects are added, strain names can be specified using the
`strains` argument:
```python3
>>> species = plot.get_species(genera=["Aspergillus"], strains=["CBS 1234", ... ])
```
Retrieve marker sequences
```python3
>>> markers = plot.get_marker_sequences(species, marker="ITS")
>>> markers
[, ... ]
```
Note this returns instances of the `Sequence` class, which have headers and sequences
```python3
>>> sequence = markers[0]
>>> sequence.header, sequence.sequence[:20]
(464, 'AAGGATCATTACCGAGTGTA')
>>> sequence.fasta()
'>464\nAAGGATCATTACCGAGTGTAGGGTTCCCTCGTGGAGCCCAACC ... '
```
Phylogenetic procedures are stored in the `phylogeny` module
```python3
from fungphy import phylogeny as phy
```
Align markers
```python3
>>> msa = phy.align_sequences(markers, name="ITS", tool="mafft")
>>> msa
```
This returns an instance of `MSA`, which is a list-like class used to store aligned
`Sequence` objects:
```python3
>>> msa.records
[, ... ]
```
Trim MSAs to first and last non gap-containing columns
```python3
>>> msa = phy.trim(msa)
```
Optionally, trim as above by a minimum number of non gap-containing columns
```python3
# Trim to first/last columns where at least 90% of sequences contain no gaps
>>> msa = phy.trim(msa, threshold=0.9)
```
Write to file in FASTA format
```python3
>>> with open("ITS.msa", "w") as fp:
... fp.write(msa.fasta())
```
Back to the `plot` module, we can automate the above steps for multiple markers
```python3
>>> mmsa = plot.align_organisms(
... species,
... markers=["ITS", "BenA", "CaM", "RPB2"],
... trim_msa=True
... )
Aligning ITS
Aligning BenA
Aligning CaM
Aligning RPB2
>>> mmsa
```
Note this returns an instance of `MultiMSA`, which is another list-like class, this time
used to store multiple instances of `MSA`
```python3
>>> mmsa.msas
[, ... ]
```
Write to file in FASTA format, as above
```python3
>>> with open("multi.msa", "w") as fp:
... fp.write(mmsa.fasta())
```
Can also report partition intervals
```python3
>>> mmsa.partitions
[(1, 617), (618, 1098), (1099, 1606), (1607, 2562)]
```
Write to file in RAxML format
```python3
>>> with open("partitions.msa", "w") as fp:
... fp.write(mmsa.raxml_partitions())
```
Ready to be analysed by e.g. modeltest-ng / raxml-ng.
MultiMSAs can also be re-read back into Python if given an accompanying RAxML format
partition file generated as above
```python3
>>> mmsa = phy.MultiMSA.from_msa_file("msa.fasta", "partitions.txt")
```
### Summary table
We can use the `Summary` class to generate a table of marker accessions for use in
publications.
```python3
>>> table = plot.Summary.make(species, markers=["ITS", "BenA", "CaM", "RPB2"])
>>> table.headers
['Organism', 'ITS', 'BenA', 'CaM', 'RPB2']
>>> table.rows
[['Aspergillus albertensis', 'EF661548', 'EF661464', 'EF661537', 'EU021628'],
['Aspergillus alliaceus', 'EF661551', 'EF661465', 'EF661534', 'MG517825'],
...]
```
Format as e.g. tab delimited file with headers
```python3
>>> formatted = table.format(delimiter="\t", show_headers=True)
>>> print(formatted)
Organism ITS BenA CaM RPB2
Aspergillus albertensis EF661548 EF661464 EF661537 EU021628
Aspergillus alliaceus EF661551 EF661465 EF661534 MG517825
Aspergillus avenaceus AF104446 FJ491481 FJ491496 JN121424
...
```
Write to file
```python3
>>> with open("markers.tsv", "w") as fp:
... fp.write(formatted)
```
### Visualisation
Note above aligned `Sequence` objects use their respective strain's database row ID as a
header. This is to allow separate `MSA` instances to be linked in multi-locus analyses.
They are also used to lookup species information from the database when reading in a
tree file for visualisation.
Generate a tree from MSA using FastTree
```python3
>>> newick = phy.fasttree(mmsa)
```
This generates a tree in Newick format, so convert to an ETE3 toolkit `Tree` object.
Note we set our outgroup as *Aspergillus avenaceus*.
```python3
>>> tree = plot.read_tree(newick, outgroup="avenaceus")
```
Visualise the tree
```python3
>>> plot.show(tree)
```
This will open the ETE3 interactive tree browser for further manipulation. Magically,
all our species information (including superscript Ts to indicate ex-type strains) have
been filled in via the database (see top of page). Also note that support values of 100
or 1.0 are shortened to asterisks (\*). Note that due to the use of HTML in the tree
leaf labels, resulting trees should be saved and edited as PDF files. Saving as SVG and
importing into e.g. InkScape will result in warped font kerning, spacing, etc.
We can also load in trees generated using e.g. raxml-ng
```python3
>>> tree = plot.read_tree_from_path("path/to/file.nw")
```
Or multiple trees for same species generated through different methods
```python3
>>> tree = plot.read_trees_from_paths(["tree1.nw", "tree2.nw"], merge=True)
```
With `merge=True`, support values for identical clades are merged; this allows
e.g. placing posterior probabilities from Bayesian inference on a maximum likelihood
tree, and vice versa. Unfortunately, since ETE3 reads in support values numerically,
it does not allow combined values generated by e.g. `IQTree -s msa --alrt 1000 -B
1000`. Thus `fungphy` does not support these either.