saezlab / plugy

Flexible processing and analysis of plug microfluidics data

Plugy: Python module for plug microfluidics data analysis

Issues

Feedback, questions, bug reports are welcome: https://github.com/saezlab/plugy/issues

Installation

Imports & Setup

You can now install plugy as a package using pip in your conda environment. To install pip in your conda environment run the following lines on your bash or conda prompt.

# Activate your conda environment replacing 'YOUR_ENVIRONMENT' with the name of your environment
conda activate YOUR_ENVIRONMENT

# Install pip git support, such that plugy can be directly installed from gitlab
conda install pip git

# Install plugy into your environment
pip install git+https://github.com/saezlab/plugy@master

# If you want to use the latest development version use this instead
pip install --force-reinstall git+https://github.com/saezlab/plugy@dev

Quick start

This notebook will show you how to run a plugy based analysis of a drug combination Braille display microfluidics experiment.

First, make sure your Python shell is running in the working directory where (or in its subdirectories) you have the data and where you want to save the results.

The simplest workflow, which is sufficient most of the times, looks like this:

import plugy
exp = plugy.PlugExperiment()
exp.main()

Further settings, parameters can be passed to PlugExperiment:

import plugy
exp = plugy.PlugExperiment(
    peak_min_threshold = 0.02,
    barcoding_param = {
        'times': (.2, 4.0),
    },
    heatmap_second_scale = 'pos-ctrl',
)
exp.main()

If you want to interact with the data use the contents of the exp object. It contains all the plug, pmt, channel and sequence data that was used in the analysis. For example, a pandas.DataFrame containing the statistics for each sample:

exp.sample_statistics

Tutorial

You can find more examples in the plugy guide: https://github.com/saezlab/plugy/blob/master/notebooks/plugy_guide.ipynb

Development history

https://github.com/saezlab/plugy/blob/master/NEWS.md

plugy's Issues

Collect examples from the issues and correspondance to one tutorial

Z-scores: by cycle or for entire screen

Give option to calculate z-score by each cycle and/or by all cycles

Move to poetry build system

Z-factor: always apply absolute value in denominator

Workflow management

Give option to do the analysis in steps (currently all at once)

We should have this already, maybe it doesn't work, have to check

Export by condition summay table

Give summarized statistics table of drug treatments as in the table below. One table that has cycles separated, and one that averages all cycles
(almost same as #5)

y-axis limits for raw PMT data

Give option to plot pmt raw data with desired y-axis range.

I believe we have this already?

Condition (drug) top list

Give a result table for ranked drug treatments, based on the chosen fold change/z-score difference (to medium control) and/or based on the chosen p-value

More plug length figures

In the QC folder, plot plug length for each drug treatment, by each cycle, and averaging all cycles

Heatmap of green signals

Provide heatmap of green signals (currently only heatmap of green/orange signals)

Config: exclude conditions / drugs

Give option to exclude certain drug treatments after the QC step

Heatmap: divergent color scale

Change heatmap colors to better distinguished between up vs down-regulation of apoptotic
signals by drug treatments vs. medium control.

• Divergent color scale
• Two stars for higher significance

Automated sphinx build

Channels → role mapping in config

Modify the configuration step such that green/orange/uv signals can be switched on demand

Alternative controls

Give option to choose one or several controls to compare drug treatments to

Plot raw green signal

In the QC folder, plot green fluorescence for each drug treatment, by each cycle, and averaging all cycles

Same for the orange channel.