Getting started

Requirements

The analysis code relies on processing on a GPU. Therefore, a graphics card with Nvidia drivers is required for running the code. We tested the code on Linux and Windows.

Get ready

Setup your Python Environment

Make sure you have Python and Anaconda or Miniconda installed on your system.

Download the code from Github

Next, please make sure you have Git already installed installed on your system. For a user-friendly experience, GitHub Desktop is a nice software tool.

Next, create a folder and go to this folder. Clone this repository to your local path, e.g. in the Anaconda Prompt.

(base) C:\Users\Username\Code> $ git clone https://github.com/DaetwylerStephan/multi-scale-image-analysis.git

Software installation

Create and activate a Python environment called imageanalysis_env

(base) MyComputer ~ $ conda create -n imageanalysis_env python=3.9 cudatoolkit=11.8.* cudnn==8.* -c anaconda

Note: This environment uses Python version 3.9 and sets the cuda tool kits for GPU-based computation.

Next, activate the environment:

(base) MyComputer ~ $ conda activate imageanalysis_env

Installation of the packages

Now navigate to the folder “multiScaleAnalysis”, where the setup.py and requirements.txt file are located, and install the packages:

(imageanalysis_env) MyComputer ~$ cd multi-scale-image-analysis
(imageanalysis_env) MyComputer ~/multiScaleAnalysis $ pip install .

Now you can start executing the scripts and test them using the testdata provided on Synapse or Zenodo (see below).

Running and editing functions with PyCharm

After establishing the Conda environment and installing all functions with “pip install .” in the Anaconda prompt, open your favorite IDE to modify and run code. Here, we use PyCharm.

First open your folder as project in PyCharm in File>Open…

Next, we want to add the environment to our project. For this go to File>Settings… and choose your recently established Conda environment.

and next

Now the Terminal is ready for input:

Before running the script, we need to indicate where the data to process is saved. Please check out the different sections for different scripts. For example, in the script to run PSF analysis, modify the parentfolder filepath.

To run the script, we need to go to the folder with the python function and enter “python PSF_measurements.py”:

(imageanalysis_env) MyComputer ~/multiScaleAnalysis $ cd Tools
(imageanalysis_env) MyComputer ~/multiScaleAnalysis $ python PSF_measurements.py

Test data location

Data to test our code is available on

Zenodo:: https://doi.org/10.5281/zenodo.12791724

The raw data for all our analysis (several TBs) is available on Synapse, the official storage of National Institute of Health, MC2 centers:

Experiment Collection:: https://doi.org/10.7303/syn61795837
Code Example Datasets Collection:: https://doi.org/10.7303/syn61795850

If you are interested in it, please check out: https://help.synapse.org/docs/Getting-Started.2055471150.html

Troubleshooting

If you run the code at an institution with a firewall, you may need to change the proxy settings to enable pip and conda to download files.

To do so, change your system environment variables (Windows). You obtain the port number (1234) and proxy address (http://proxy.your_university.edu) from your system administrators.

Variable = HTTP_PROXY; Value = http://proxy.your_university.edu:1234

Variable = HTTPS_PROXY; Value = https://proxy.your_university.edu:1234

If you continue to have issues then change the value of Variable HTTPS_PROXY to http://proxy.your_university.edu:1234

If this does not resolve your download/proxy issues, also update the configuration files for conda and pip to include the proxy settings. For Windows, the paths are saved at:

The conda configuration file can be found at C:\Users\UserProfile\.condarc

The pip configuration file can be found at C:\Users\UserProfile\pip\pip.ini