Features

PyCAT is a comprehensive Python toolkit for analyzing biomolecular condensates and bioimages. It offers robust analysis methods, customizable pipelines, and visualization tools for in-depth biological insights. This page is a reference for features and tools available in PyCAT.

Analysis Methods Available

PyCAT offers a wide range of analysis methods for quantifying bio-images.

Region Properties Analysis : Measure area, intensity, shape, texture, and other features of ROIs (regions of interest) like condensates, cells, etc.

Object-Based Colocalization Analysis : Object ROI mask comparisons for segmented object masks

• Mander’s M1 value

• Mander’s M2 value

• Jaccard Index

• Sorensen-Dice Coefficient

• Calculate Distance Between Objects

• Percent of Objects Non-coincident

• Mander’s Colocalization Coefficient

Pixel-Wise Correlation Analysis : Correlation Analysis for images without distinctive/segmentable objects

• Pearson’s R value

• Spearman’s R value

• Kendall’s Tau value

• Weighted Tau value

• Li’s ICQ value

• Mander’s Overlap Coefficient

• Mander’s k1 value

• Mander’s k2 value

Modified Costes Analysis : Automatically generated thresholds and statistical significance testing for correlation analyses

• Costes Automatic Thresholded M1 M2

• Calculate Costes Significance

• Perform Modified Costes Thresholding

Correlation Function Analysis : Auto- and Cross-Correlation Functions with Gaussian fits

• Fitted gaussian parameters (mu, sigma, etc)

Analysis Pipelines

PyCAT also provides several pre-configured analysis pipelines for common use cases.

Condensate Analysis : Tailored analysis for in-cellulo condensates

Colocalization Analysis : Object-based and pixel-wise correlation pipelines

General ROI Analysis : Exploratory pipeline with region property measurements

Fibril Analysis : Analyze beta-amyloid fibers and fibril structures

Toolbox Reference

This reference table lists all of the individual functions available in the PyCAT toolbox along with brief descriptions.

Toolbox Functions

Function Category	Description
Image Processing Tools	Tools for image pre-processing, enhancement, and noise reduction.
Pre-processing	Applies top hat, LoG, and other image processing filters.
Rescale Intensity	Rescales image intensity values to a desired range.
Invert Intensity	Inverts pixel intensity values (e.g., dark to bright).
Upscale Image	Increases image resolution while preserving structural features.
RB Gauss BG removal	Rolling ball and Gaussian background removal.
BG removal with edge enhancement	Removes background while enhancing peaks and edges through Gabor filtering.
WBNS	Wavelet Background and Noise Subtraction.
Wavelet noise reduction	Reduces noise using wavelet transforms for multi-scale denoising.
Bilateral noise reduction	Smoothens images while preserving edges through bilateral filtering.
CLAHE	Contrast-Limited Adaptive Histogram Equalization for local contrast enhancement.
Peak and Edge enhancement	Enhances image peaks and edges via Gabor and LoG filtering.
Morphological Gaussian filter	Smoothens images while applying morphological opening and Gaussian filtering.
LoG enhancement	Laplacian-of-Gaussian enhancement for edge detection.
Deblur by Pixel Reassignment	An advanced PSF and deblurring tool which upscales images in the process.

Image Segmentation Tools	Tools for segmenting images into meaningful regions of interest (ROIs).
Local Thresholding functions	Performs segmentation using local thresholding algorithms.
Cellpose segmentation	Uses Cellpose, a deep learning-based method for cell segmentation.
Random Forest classifier segmentation	Segments images using a Random Forest classifier model.
Felzenszwalb segmentation and RAG region merging	Segments images using graph-based Felzenszwalb algorithms and merges regions.

Label and Mask Tools	Tools for working with binary masks and labeled mask regions.
Binary Mask Morphological Operations	Applies morphological operations (e.g., dilation, erosion) to binary masks.
Measure Binary Mask	Extracts measurements (area, intensity) from binary masks.
Label Binary Mask	Assigns unique labels to connected components in binary masks.
Convert Labels to Mask	Converts labeled regions into binary masks.
Update Labels	Updates or modifies labels in labeled images.
Region Props on Labeled Mask	Computes region properties (e.g., area, eccentricity) for labeled regions.

Layer Operations	Tools for managing and merging image layers.
Simple Multi-layer merging	Merges multiple image layers into a single layer.
Advanced 2-layer merging	Performs advanced merging operations between two image layers.

Colocalization and Correlation Tools	Tools for analyzing colocalization and correlation between regions or channels.
Object-Based Colocalization	Performs object-based analysis with metrics like Manders, Sorensen-Dice, and object distances.
Pixel-Wise Correlation	Computes pixel-wise correlations using Pearson, Weighted Tau, Li’s ICA, and other metrics.
Correlation Function (ACF, CCF)	Calculates Auto-Correlation (ACF) and Cross-Correlation Functions (CCF).

Data Visualization	Tools for visualizing analysis results.
Plotting Widget	Interactive widget for plotting and visualizing analysis outputs.

Cell Analysis Features

This table lists the features measured in the cell_df data structure when using the condensate analysis pipeline and the cell analyzer and condensate analyzer functions.

Features Measured in cell_df

Feature	Description
label	Unique identifier for each object (cell or ROI).
area	Total area of the object in pixels.
intensity_mean	Mean intensity of the object.
axis_major_length	Length of the major axis of the object’s fitted ellipse.
axis_minor_length	Length of the minor axis of the object’s fitted ellipse.
eccentricity	Deviation from circularity; 0 for a circle, 1 for a line.
perimeter	Perimeter length of the object in pixels.
intensity_std_dev	Standard deviation of the object’s intensity.
intensity_median	Median intensity value of the object.
intensity_total	Total summed intensity of the object.
cell_micron_area	Object area in square microns, based on image resolution.
image_resolution_um_per_px_sq	Image resolution in (um/px)^2
cell_snr	Signal-to-noise ratio: mean cell intensity / std dev of background (non-cell) intensity.
gaussian_snr_estimate	Gaussian SNR: mean cell intensity / Gaussian background noise estimate.
contrast	Contrast measure for the object region.
dissimilarity	Texture-based measure of intensity dissimilarity.
homogeneity	Texture-based measure of regional intensity uniformity.
ASM	Angular Second Moment; measures image uniformity.
energy	Energy metric derived from the image region.
correlation	Correlation measure between neighboring pixel intensities.
32_bit_entropy	Entropy of the object calculated from 32-bit float intensities.
8_bit_entropy	Entropy of the object calculated from 8-bit unsigned integer intensities.
8_bit_entropy_img_avg	Average entropy of the 8-bit image.
img_kurtosis	Measure of the “tailedness” of the intensity distribution.
standardized_sixth_moment	Standardized sixth moment of the intensity distribution.
kurtosis_z_score	Z-score for kurtosis, indicating deviation from normality.
p_val	p-value for statistical significance of kurtosis.
lbp_mean	Mean of the local binary pattern (LBP) features.
lbp_std	Standard deviation of LBP features.
lbp_entropy	Entropy of the LBP features.
puncta_micron_area_mean	Mean size of puncta within a cell, in square microns.
puncta_micron_area_std	Standard deviation of puncta sizes within a cell, in square microns.
puncta_ellipticity_mean	Mean ellipticity of all puncta within a cell.
puncta_intensity_total	Total intensity: mean intensity * area of all puncta.
puncta_intensity_dist_mean	Mean intensity of all puncta within the cell.
number_of_puncta	Total number of puncta detected within the cell.
cell_xor_puncta_int_mean	Mean intensity of the region inside the cell, excluding puncta regions.
cell_xor_puncta_int_std	Standard deviation of intensity in the cell region, excluding puncta ROIs.
cell_xor_puncta_int_total	Total intensity of the region inside the cell, excluding puncta ROIs.
cell_xor_puncta_area	Area of the region inside the cell, excluding puncta ROIs (square microns).
snr_test	Signal-to-noise ratio: mean puncta intensity / std dev of dilute phase intensity.
partition_test	Partition coefficient: puncta intensity mean / cell XOR puncta intensity mean.
partition_test_total_int	Partition coefficient: puncta total intensity / cell XOR puncta total intensity.
spark_score	Total puncta intensity / total cell intensity.
puncta_classifier	Binary classifier for puncta presence: 1 (puncta), 0 (none).

Puncta Analysis Features

This table lists the features measured in the puncta_df data structure when using the condensate analysis pipeline and the condensate analyzer function.

Features Measured in puncta_df

Feature	Description
label	Unique identifier for each punctum (object).
area	Total area of the punctum in pixels.
intensity_mean	Mean intensity of the punctum.
axis_major_length	Length of the major axis of the punctum’s fitted ellipse.
axis_minor_length	Length of the minor axis of the punctum’s fitted ellipse.
eccentricity	Deviation from circularity; 0 for a circle, 1 for a line.
perimeter	Perimeter length of the punctum in pixels.
ellipticity	Measure of elongation: 1 - axis_minor_length / axis_major_length. Higher values indicate more elongated puncta.
circularity	Measure of shape compactness, normalized for irregular shapes: 4 * pi * area / (perimeter ** 2). Note: Normalized to account for the “coast of England” fractal paradox, where irregular shapes cause perimeter to scale non-linearly with area.
micron area	Area of the punctum in square microns, based on image resolution.
cell label	Label of the corresponding cell to which the punctum belongs.