Skip to main content

Rapid transformations for cytometry data

Project description

Cytotransfrom

PyPI PyPI - Python Version PyPI - Wheel PyPI - License Codecov GitHub Workflow Status (with event)

Description

Cytotransform is a python package for transforming flow cytometry data. It implements the following transformations according to the GatingML 2.0 definitions (https://flowcyt.sourceforge.net/gating/latest.pdf):

  • Parametrized logarithmic transformation
  • Parametrized inverse hyperbolic sine transformation (asinh)
  • Logicle transformation
  • Hyperlog transformation

Each transformation is implemented as a Transform class with a transform method that takes a numpy array as input and returns the transformed array. The Transform class also has a transform_inverse method that takes a numpy array as input and returns the inverse transformed array. Each implementation includes validation of the input parameters. The transform classes support multiprocessing out of the box and if n_jobs is set to more than 0, then the input data will be split into n batches depending on the number of cores available and each batch will be transformed in parallel. If n_jobs is set to -1, then all available cores will be used. If n_jobs is set to 0, then no multiprocessing will be used.

Cytotransform is thanks to the fantastic community of scientists and developers in the single cell and flow cytometry data analysis ecosystem. It implements the FastLogicle C++ library for logicle transformations originally implemented by Wayne A Moore and David R Parks (see https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4761345/). Code was inspired by the great work by Scott White (https://github.com/whitews/FlowKit) and Brian Teague (https://github.com/cytoflow).

Installation

Cytotransform can be installed from PyPi using pip:

pip install cytotransform

Usage

The transform and inverse_transform methods take a numpy array or Pandas DataFrame as input and return a numpy the transformed array/dataframe.

Parametrized logarithmic transformation

from cytotransform import ParametrizedLogTransform
transformer = ParametrizedLogTransform(m=4.5, t=262144, n_jobs=-1)
transformed_data = transformer.transform(data)
data = transformer.inverse_transform(transformed_data)

Parametrized inverse hyperbolic sine transformation (asinh)

from cytotransform import AsinhTransform
transformer = AsinhTransform(m=4.5, t=262144, a=0.0, n_jobs=-1)
transformed_data = transformer.transform(data)
data = transformer.inverse_transform(transformed_data)

Logicle transformation

from cytotransform import LogicleTransform
transformer = LogicleTransform(t=262144, w=0.5, m=4.5, a=0.0, n_jobs=-1)
transformed_data = transformer.transform(data)
data = transformer.inverse_transform(transformed_data)

Hyperlog transformation

from cytotransform import HyperlogTransform
transformer = HyperlogTransform(t=262144, w=0.5, m=4.5, a=0.0, n_jobs=-1)
transformed_data = transformer.transform(data)
data = transformer.inverse_transform(transformed_data)

License

Cytotransform is licensed under the MIT license, is free to use, and comes with no warranty whatsoever.

Project details


Download files

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

Source Distribution

cytotransform-0.2.1.tar.gz (14.3 kB view hashes)

Uploaded Source

Built Distribution

cytotransform-0.2.1-cp311-cp311-manylinux_2_35_x86_64.whl (1.1 MB view hashes)

Uploaded CPython 3.11 manylinux: glibc 2.35+ x86-64

Supported by

AWS AWS Cloud computing and Security Sponsor Datadog Datadog Monitoring Fastly Fastly CDN Google Google Download Analytics Microsoft Microsoft PSF Sponsor Pingdom Pingdom Monitoring Sentry Sentry Error logging StatusPage StatusPage Status page