Quantifying Colocalization: Thresholding, Void Voxels and the Hcoef

Abstract

A critical step in the analysis of images is identifying the area of interest e.g. nuclei. When the nuclei are brighter than the remainder of the image an intensity can be chosen to identify the nuclei. Intensity thresholding is complicated by variations in the intensity of individual nuclei and their intensity relative to their surroundings. To compensate thresholds can be based on local rather than global intensities. By testing local thresholding methods we found that the local mean performed poorly while the Phansalkar method and a new method based on identifying the local background were superior. A new colocalization coefficient, the Hcoef, highlights a number of controversial issues. (i) Are molecular interactions measurable (ii) whether to include voxels without fluorophores in calculations, and (iii) the meaning of negative correlations. Negative correlations can arise biologically (a) because the two fluorophores are in different places or (b) when high intensities of one fluorophore coincide with low intensities of a second. The cases are distinct and we argue that it is only relevant to measure correlation using pixels that contain both fluorophores and, when the fluorophores are in different places, to just report the lack of co-occurrence and omit these uninformative negative correlation. The Hcoef could report molecular interactions in a homogenous medium. But biology is not homogenous and distributions also reflect physico-chemical properties, targeted delivery and retention. The Hcoef actually measures a mix of correlation and co-occurrence, which makes its interpretation problematic and in the absence of a convincing demonstration we advise caution, favouring separate measurements of correlation and of co-occurrence.

Quantifying colocalization: The case for discarding the Manders overlap coefficient

“…Our current simulations, previous work and the equation itself shows that a zero value only occurs when the two molecules never co-occur. This is quite different from a perfect negative correlation where high values for one would be associated with low values for the second and vice versa, together with an intermediate population with more similar values [5]. We have previously demonstrated that populations with perfect negative correlations can have values at the high end of the MOC's range [3,13] and here demonstrate that even a perfect positive correlation plus complete co-occurrence does not always produce a maximal MOC.…”

“…Some coefficients fall outside our bipolar scheme and instead merge the two primary attributes into a single value. These we categorize as hybrids which we suggest includes the Manders Overlap Coefficient (MOC) [2,3] and the H coeff [4,5].…”

EzColocalization: An ImageJ plugin for visualizing and measuring colocalization in cells and organisms

“…Several factors limit the use of current software for visualizing the localization of reporters in biological samples and measuring colocalization 9 – 12 . One factor is that customization of the software is often required for the equipment, reporters and samples 13 , 14 , and for automated analyses.…”

MatCol: a tool to measure fluorescence signal colocalisation in biological systems

“…However, the inability to automatically select ROIs makes analysis of a large number of images extremely difficult 10 . Background can vary significantly in different portions of the images; therefore, global thresholding to remove background is not a suitable option as discussed by Regeling et al ., as well as Adler and Parmryd 11 , 12 . Different coefficients report colocalisation using a range of 0 to 1 or −1 to 1, which may lead to investigators considering different coefficients significant 8 , 13 .…”