Automated gold particle quantification of immunogold labeled micrographs

Abstract

Background: Immunogold cytochemistry is the method of choice for precise localization of antigens on a subcellular scale. The process of immunogold quantification in electron micrographs is laborious, especially for proteins with a dense distribution pattern. New methods: Here I present a MATLAB based toolbox that is optimized for a typical immunogold analysis workflow. It combines automatic detection of gold particles through a multi-threshold algorithm with manual segmentation of cell membranes and regions of interests. Results: The automated particle detection algorithm was applied to a typical immunogold dataset of neural tissue, and was able to detect particles with a high degree of precision. Without manual correction, the algorithm detected 97% of all gold particles, with merely a 0.1% false-positive rate. Comparisons with existing method(s): To my knowledge, this is the first free and publicly available software custom made for immunogold analyses. The proposed particle detection method compares favorably to previously published algorithms. Conclusions: The software presented here will be valuable tool for researchers in neuroscience working with immunogold cytochemistry.