Challenges with Integration of Third-Party libraries in Fourier Shell Correlation Ananlysis

[aadedolapo]

1. Integration of toupy codes into FSC.py:

While attempting to integrate code segments from the toupy package source code from their repo into the FSC.py file, an obstacle arose during the incorporation of the fastfftn library. This led to a roadblock in the code development process.

2. Adding toupy as a dependency in setup.cfg

A search for libraries capable of conducting fast 3D Fourier Transform operations was conducted. Unfortunately, no suitable libraries were discovered, prompting the decision to incorporate the toupy package along with its dependencies into the CoMeTr package. Subsequent efforts encountered an error due to the usage of the skimage.feature.register_translation method, which has been deprecated within the skimage library. This has hindered progress on the FSC_toupy.py file.

3. Quoll and Miplib Library Integration

As a further avenue of exploration, the quoll and miplib libraries were considered because they both have available methods to calculate the Fourier Ring Correlation and Fourier Shell Correlation. Attempting to utilize the quoll library for Fourier Ring Correlation (FRC) analysis faced constraints as the library only supports 2-dimensional images. This limitation necessitated the slicing of data at a specific z-index, preserving information on the x and y axes while reducing the analysis to two dimensions. However, this approach has encountered a deadlock at line 71 on the FSC_quoll.py file. An effort to utilize the quoll.reader.Image to load 2D arrays as quoll images achieved success, but subsequent stages remained inaccessible. The attempt to instantiate a miplibImage object resulted in a distinct error message, further complicating the integration of the miplib library into the workflow.