SAFT : Semi Automatic Fluorescent Trace analysis

v. 0.5 2022-08-22

SAFT was written in Python 3 for macOS. In principle it is cross-platform, in practice it was never launched on Windows or Linux. If you happen to get it working on either of these platforms, please let us know. SAFT runs normally on M1 Macs.

This software package is described in the following publication: Yuchen Hao, Estelle Toulmé, Benjamin König, Christian Rosenmund, Andrew J R Plested (2023) Targeted sensors for glutamatergic neurotransmission eLife 12:e84029 https://doi.org/10.7554/eLife.84029

Getting Started on macOS (for Python/UNIX beginners)

(Python/UNIX experts see below)

Note, we are sorry, the following isn't trivial. We do work on packaging SAFT into a clickable app, it's just not easy either.

1. Clone the SAFT repository to your machine by clicking the green "Code" button at the top of this page and selecting "Download Zip".

You will receive a directory called "SAFT-master", probably in the 'Downloads' folder.

2. To get all the Python packages you need to run SAFT, we recommend using miniconda.
   Download miniconda for python 3.6 bash ; we provide everything you need after that.
   Here's the link. https://docs.conda.io/projects/continuumio-conda/en/latest/user-guide/install/macos.html

3. Use Spotlight (Command+Space) to find and launch 'Terminal'. In the following, you will need to enter commands in the Terminal. You must be exact...

4. To install miniconda, you need to navigate to the directory where it downloaded, probably again 'Downloads'.

cd Downloads

5. Type the following command to make the installation. Use the space bar to page through the licence agreement. Answer (type) 'yes' to everything...

bash Miniconda3-latest-MacOSX-x86_64.sh

(Free UNIX tip: if you start typing bash Min and then press the Tab key "->|", it will autocomplete)

6. After this you'll need to restart the Terminal window (command + W, then command + N)

7. Now you need to navigate to the SAFT directory in the Terminal. You start in your "Home" folder (its name is your username).
   The command you need is cd (change directory):

cd Downloads\SAFT-master

8. To prepare for SAFT, get Conda to create the SAFT environment:

conda env create --file SAFTenv.yml

9. Once all the packages have downloaded and Conda has arranged them all, to use them you need to switch to this environment:

conda activate SAFT

10. To run SAFT, type the following command into the Terminal:

pythonw SAFT.py

On the first run in a freshly installed environment, you will have to wait for some of the packages. Matplotlib is kind enough to note that it is spending the time installing fonts.

From the window, you can open the example file SAFT_Data.xlsx for analysis. Read instructions.pdf for how some information about how to use the program.

Note: You can also launch quantal fitting of histograms as a standalone app. If you add test some example data (the file ExPeak_Data.xlsx) will load automatically.

pythonw fitHistograms.py pythonw fitHistograms.py test

Experts

---

SAFT depends on quite a few packages, for its GUI (Qt, pyqtgraph, pyside2, matplotlib, pyobjc-framework-Cocoa), for number crunching (numpy, scipy, pandas) and for data handling (pandas, xlrd, openpyxl). Getting all these packages aligned in mutually compatible versions is not easy, even with package management. You can use anaconda, pip or compile from source. In our experience, it is best to start with a fresh environment.

Conda currently provides Qt 5.9.7 and this makes it convenient to use conda to set up a fresh environment. To make getting the right environment in place easy, we include a conda environment .yml file in the distribution (SAFTenv.yml). To use this, you must have Anaconda or Miniconda installed on your machine. If you don't have Anaconda already, just get Miniconda: https://docs.conda.io/projects/continuumio-conda/en/latest/user-guide/install/macos.html.

It is not recommended to work with any of the versions of Python that Apple bundles with OSX. Please have conda install Python 3.7.9 (if you don't have it) before doing anything further.

To install the packages needed for SAFT into a fresh environment with Miniconda, go to the SAFT directory and enter the following in the Terminal,

conda env create --file SAFTenv.yml

This will create an environment called 'SAFT' with a working set of packages.

As conda will tell you, to switch to this environment, enter the following,

conda activate SAFT

then navigate to the SAFT directory and execute the commands as above to launch SAFT.

If you happen to already have an environment named SAFT, you would need to remove it first:

conda env remove --name SAFT

SAFT runs normally on M1 Mac, if you have upgraded your machine since installing, issue the following command to get packages up to date:

conda env update --file SAFTenv.yml

Very important : Qt versions after 5.9 are not open source (breaking the conditions of our license), and coincidentally have serious bugs that break the GUI on macOS. We need Qt for the high-performance scientific plotting library pyqtgraph: http://www.pyqtgraph.org.
PySide2 versions up to 5.15 seem to run happily with Qt 5.9.7 but it may be difficult to avoid a concurrent Qt update, unless you have a separate version of Qt compiled from open source and know how to point PySide2 at it. The latter solution is not stable.

Dependencies as of 2020-11-12

---

Python 3.7.9

qt==5.9.7
PySide2>=5.9.0a1
numpy==1.19.2
scipy==1.5.2
pyqtgraph==0.11.0
pandas==0.24.0
matplotlib==3.3.2
xlrd==1.2.0
openpyxl==3.0.5
pyobjc-framework-Cocoa==6.2.2
python.app 2