CellScanner

CellScanner is a tool that counts different microbial species in flow cytometry data of communities separately with the help of a classifier that is trained on mono-cultures. It is implemented in Python and compatible with Windows, Mac and Linux.

Version 2

Version 2 is a complete re-implementation of CellScanner with a new (simplified) interface that addresses several of the shortcomings of version 1, such as speed, support for dead/live staining and re-use of a trained model on several co-cultures. It makes use of several improvements introduced by Master student Vladislav Vlasov and incorporates a preprocessing step with UMAP. The UMAP step is based on a script by Daniel Garza (available here) that he developed for this ISME publication. Version 2 was implemented by Master student Ermis Delopoulos and additional functions were added by Haris Zafeiropoulos.

Installation:

git clone https://github.com/msysbio/CellScanner.git
cd CellScanner
conda create -n cellscanner python=3.12.2 
conda activate cellscanner
pip install -r requirements.txt

More detailed installation instructions and an exe for windows are available here.

Help:
Documentation and a tutorial are available here.

Code:
The Github repository for version 2 is available here.

Differences to version 1:
The differences with version 1 are documented here.

Validation:
Validation on a previous data set with known ratios available here.

Version 1

Version 1 offers three classifiers, namely neural networks, random forests and logistic regression. It also supports unsupervised clustering and includes a number of gut bacterial mono-culture and control flow cytometry files. Version 1 was developed by PhD student Clémence Joseph, who also wrote the manual and performed the evaluations.

Manual:
The manual is available here.

Validation:
Version 1 was validated in silico and in vitro on gut bacterial flow cytometry data as described here. More details on implementation and a more comprehensive evaluation are available here (figures, tables and supplementary material included at the end of the document).

Download:
For Windows, an executable is available. For Mac and Linux, please follow the instructions from the INSTALL file.
Windows: CellScanner1.0.1.exe. Please execute as an administrator
Mac/Linux: CellScanner.zip.
The Github repository for version 1 is available here.