Abstract:BackgroundFluorescence microscopy is widely used to determine the subcellular location of proteins. Efforts to determine location on a proteome-wide basis create a need for automated methods to analyze the resulting images. Over the past ten years, the feasibility of using machine learning methods to recognize all major subcellular location patterns has been convincingly demonstrated, using diverse feature sets and classifiers. On a well-studied data set of 2D HeLa single-cell images, the best performance to date, 91.5%, was obtained by including a set of multiresolution features. This demonstrates the value of multiresolution approaches to this important problem.ResultsWe report here a novel approach for the classification of subcellular location patterns by classifying in multiresolution subspaces. Our system is able to work with any feature set and any classifier. It consists of multiresolution (MR) decomposition, followed by feature computation and classification in each MR subspace, yielding local decisions that are then combined into a global decision. With 26 texture features alone and a neural network classifier, we obtained an increase in accuracy on the 2D HeLa data set to 95.3%.ConclusionWe demonstrate that the space-frequency localized information in the multiresolution subspaces adds significantly to the discriminative power of the system. Moreover, we show that a vastly reduced set of features is sufficient, consisting of our novel modified Haralick texture features. Our proposed system is general, allowing for any combinations of sets of features and any combination of classifiers.

Self organizing maps for class discovery in the quantitative colocalization analysis feature space

A correlation analysis framework for localization-based super-resolution microscopy

Correlation Analysis Framework for Localization-Based Superresolution Microscopy.

Learning Consistent Subcellular Landmarks to Quantify Changes in Multiplexed Protein Maps.

Self organising maps for visualising and modelling

An automatic tool to analyze and cluster macromolecular conformations based on self-organizing maps

Improved Recognition of Figures Containing Fluorescence Microscope Images in Online Journal Articles Using Graphical Models.

Visualisation Of Distributions And Clusters Using Visoms On Gene Expression Data

Improving Self-Organizing Maps with Unsupervised Feature Extraction

Self-organizing feature map for cluster analysis in multi-disease diagnosis

AMSOM: Adaptive Moving Self-organizing Map for Clustering and Visualization

ViSOM - a novel method for multivariate data projection and structure visualization.

Self-organizing subspace clustering for high-dimensional and multi-view data

Exploratory analysis of excitation–emission matrix fluorescence spectra with self-organizing maps—A tutorial

Neural maps in remote sensing image analysis

Deep embedded self-organizing maps for joint representation learning and topology-preserving clustering

Bayesian machine learning analysis of single-molecule fluorescence colocalization images

A multiresolution approach to automated classification of protein subcellular location images

Automatic Detection of Subcellular Particles in Fluorescence Microscopy Via Feature Clustering and Bayesian Analysis

Quantitative Colocalization Analysis of Multicolor Confocal Immunofluorescence Microscopy Images: Pushing Pixels to Explore Biological Phenomena

Interactive and semantic data visualisation using self-organising maps