View on GitHub

FRETzel

Analyse FRET signals in single cells

User guide

Introduction

FRETzel performs cell segmentation and FRET analysis of cells over a time series in a MATLAB guided user interface (GUI). Useful for identifying cells from brightfield images in messy and crowded data sets, such as those produced from the growth of adipocytes (fat cells), but can also be used on other cells, such as yeast.

See diplayGUI.jpg and completeGUI.jpg for the user interface at opening and once an experiment using adipocytes has been analysed.

System requirements

Executables require MATLAB runtime environments as detailed in their folders. Tested on Windows 10 and Mac OSX. Typical install time is a few minutes.

Using the GUI

If running the source code in MATLAB, ensure all the code is downloaded (see Getting started) and added to MATLAB path using ‘set path’. The process of the GUI is as follows:

1. Loading files

2. Cell selection, segmentation and analysis

Irregular cells
Analysis

3. Saving files

Calculating FRET data

The software uses the time stamps from each of the .lsm files in the user selected folder to order the data. The final cell outlines are used for each of the fluorescence channels in each file to plot intensity of each fluorescence (FRET, donor and acceptor) and the FRET ratio (FRET fluorescence/donor fluorescence).

Parameters

See top right corner of displayGUI.jpg for parameters needed to run the GUI. Descriptions of each are below.

File type

Preset to ‘lsm’, can be any of the Bio-formats file formats. Must be input without a dot infront (‘lsm’ not ‘.lsm’).

Approximate diameter of cells

This must be input in microns. It is preset to 20 microns. It is used to calculate the size of the ellipses created in the cell segmentation process, based on the pixel size of the image.

Active contour number of iterations

This is a number between 0 and 100, preset to 50 that determines the number of iterations MATLAB function activecontour carries out. activecontour segments the cells once the user has confirmed the ellipses to be the desired size and shape. For messier data sets (i.e those using adipocytes or that have noise) a lower number of iterations (around 20) is more useful for identifying cell outlines. For data sets with low noise and more uniform cells (i.e. yeast cells) a higher number of iterations (70-100) is more useful. This number can be changed to find the optimum for each data set.

Number of channels

The number of channels is the total number of channels in each file, which can be more than the 4 specified in ‘File requirements’ above, but needs to be specified for the analysis of the cells across all files. Preset to 4.

Input time interval or read time interval from files

There is the option for the user to set a uniform time interval between channels (in seconds), or to tick the box ‘Read time interval from files’ to get the time data from the metaData of each file. One of these options must be input to plot the donor, FRET and acceptor graphs with time.

If the user inputs a time interval, then it is important that the files in the folder have uniform names that have consecutive numbering in order of when they were aquired. For example - control01.lsm, control02.lsm, …, control10.lsm, control11.lsm (but not control1.lsm, control2.lsm, …, control10.lsm, control11.lsm).

First frame is a ‘pre’ frame

If this box is ticked, the GUI will set the first file in the time series to time = -5 minutes, assuming it is a ‘pre’ frame (i.e. before the addition of of any solution used in the time series to measure the change in FRET). It sets the second file to time = 0 minutes. If it is not ticked, the GUI will set the first file to time = 0 minutes. In both cases each consecutive file is set using the time interval specified or found in the metaData.

File requirements

Software can use any Bio-Formats file formats (i.e. .tiff, .tif, .lsm). Channels within the files must be in a certain order: donor fluorescence, FRET fluorescence, acceptor fluorescence then a brightfield/scattered light image. There can be more than these 4 channels but the FRET and brightfield channels must be the first 4 and in the correct order. For each time series the files must be in a separate folder.

The GUI can process four different types and combinations of time series:

  1. Separate files for each time frame of the time series (for example, control01.lsm, control02.lsm, control03.lsm… in a single folder). Where there is a pre frame.
  2. Separate files for each time frame of the time series (for example, control01.lsm, control02.lsm, control03.lsm… in a single folder). Where there is not a pre frame.
  3. A time series combined in one file (for example, a folder that contains control01.lsm, that has the time frames and respective channels in one file).
  4. A separate pre frame file and a time series combined in one file (for example, a folder that contains control01.lsm as the pre frame and control02.lsm, that has the remaining time frames and respective channels in one file).

A pre frame is a frame taken before the addition of of any solution used in the time series to measure the change in FRET, it must have the same field of view as the following files in the time series. The GUI can differentiate between these 4 combinations using the user set parameters (see above).

The GUI can be slower with files 50000KB and above, but still works.

Information about downloads required

FRETzel.m

Code used to run the GUI, made in MATLAB’s GUIDE. Loads Bio-formats files using ImEx1.m, segments cells using imellipse, createMask and activecontour and runs the analysis of cells, calculating the FRET ratio and plotting graphs.

GUIDE will be removed in a future release of MATLAB. The software will still be usable but not editable. At some point the GUI will be migrated or made an executable file.

FRETzel.fig

The GUI window.

bfmatlab

MATLAB Bio-Formats toolbox from Open Microscopy Environment.

ImEx1.m

Uses Bio-Formats toolbox to open and access data from microscopy files.