Epitope aligner

Easily map epitope coordinates between sequences in an alignment, regardless of which coordinate system you are using. This lets you combine epitopes from different sources and calculate things like epitope density in a set of proteins.

epitope_aligner is a python package hosted on Github at BarinthusBio/epitope_aligner.

Full documentation at barinthusbio.github.io/epitope_aligner.

If you have any suggestions or problems, please open an issue.

Contents

• Install
• Examples
  • Minimal
  • Quickstart
  • Cookbook
• Docs
• Dev

Install

Install from pip with:

pip install epitope-aligner

Examples

Minimal examples

In the current minimal example we'll:

• convert epitope coordinates to an aligned antigen
• float the epitope sequences to match it
• calculate the number of epitopes at each position in the antigen

For the inverse of these aligning and floating operations see the cookbook.

Import functions from epitope_aligner modules and pandas to create an example dataset.

from epitope_aligner import map, stretch, utils
import pandas as pd

We'll define a short example antigen sequence, with an aligned and unaligned version.

aligned_seq = "ABC---DEFGH-IJK--LM"
seq = aligned_seq.replace("-","")

We'll define some exmple epitopes with positions in the unaligned antigen sequence.

epitopes = pd.DataFrame({
        'start':  [2,      6,      9],
        'end':    [4,      9,      12],
        'seq':    ["BCD",  "FGHI", "IJKL"],
        "length": [3,       4,     4]
})
epitopes

#    start  end   seq  length
# 0      2    4   BCD       3
# 1      6    9  FGHI       4
# 2      9   12  IJKL       4

Let's calculate the start positions of these epitopes in the aligned antigen sequence.

epitopes['newstart'] = map.align_coords(
    table = epitopes,
    aligned_parent_seq = aligned_seq,
    coord_col = "start",
    index = 1
)
epitopes

#    start  end   seq  length  newstart
# 0      2    4   BCD       3         2
# 1      6    9  FGHI       4         9
# 2      9   12  IJKL       4        13

Now we can "float" an epitope to line up with its antigen based on a start position and antigen sequence.

epitopes['float'] = map.float_epitopes(
    table=epitopes,
    parent_seq=aligned_seq,
    start_col="newstart",
    index=1,
)
epitopes

# Aligned antigen
# ABC---DEFGH-IJK--LM

# Aligned epitopes
# -BC---D
# --------FGH-I
# ------------IJK--L

We can easily count the number of epitopes overlapping each position by "stretching" them. For plotting, it is often helpful to add zeros for positions with no epitopes.

stretched_epitopes = stretch.stretch(epitopes)
positional_count = stretched_epitopes.groupby("position").size()
positional_count = stretch.add_empty_positions(
    positional_count,
    parent_seq_length=len(seq),
    index=1,
    empty_value=0
)
positional_count

# position
# 1     0.0
# 2     1.0
# 3     1.0
# 4     1.0
# 5     0.0
# 6     1.0
# 7     1.0
# 8     1.0
# 9     2.0
# 10    1.0
# 11    1.0
# 12    1.0
# 13    0.0
# dtype: float64

Read the cookbook for tips on calculating more interesting measures than counts.

Quickstart

A real world example is demonstrated in the quickstart which analyses and plots the epitopes from different strains of the influenza virus.

Cookbook

The cookbook provides a detailed description and example of all functions.

Docs

The full documentation includes function APIs under the submodules:

• map
• stretch
• utils

Dev

Details on testing, creating docs, and virtual envinments.

Dev: Set up

Create a virtual environment with python3 -m venv .venv. Activate that environment with . .venv/bin/activate. Install in editable mode with pip install -e .. Deactivate it with deactivate.

Dev: Nox

Linting, bandit, documentation, examples, and testing can all be run with nox based on noxfile.py. This is also run by github actions.

Dev: Make docs

The full guide is docs/README.md but in short pdoc generates the api documentation and renders the read me, jupyter notebook examples are converted to html, and the complete docs are hosted at barinthusbio.github.io/epitope_aligner/index.html.

Generating the docs and hosting them is handled by the github actions, but if you want to produce them locally just run nox.

Dev: Publish to PyPI

Uploading requires the build and twine packages, pip install --upgrade twine build.

python -m build will create both the --sdist and --wheel. twine check dist/* will check the package is ready for uploading. twine upload dist/* will actually upload to pypi.