An interactive tool to analyze data from protein-ligand binding experiments

Interactive Matlab tool to analyze phosphorimager scans from DRaCALA experiments. Differential Radial Capillary Action of Ligand Assay (DRaCALA) is a biochemical technique to measure the binding of a protein (purified or from cell lysate) to a radio-labelled small molecule. Data are in the form of small "spots" of protein or lysate with the labelled molecule. Strong binding inhibits the ability for the small molecule to diffuse throughout the spot, and produces a localized signal in the form of a smaller inner "spot" within the larger spot.

QuantDRaCALA provides an intuitive interface with a simple design that allows the user to load a phosphorimager-scanned file from their DRaCALA experiment, very quickly identify and pre-process each spot, create inner and outer circles over these spots, then analyze and save their data. Images with QL-value pixels are automatically converted to PSL-values, so the user can directly load their raw scanned images (although directly loading .img/.inf files is still in-progress). This tool was designed for more high-throughput experiments, where DRaCALA is performed in a 96-well plate. Manually analyzing these data can be time-consuming and repetitive (as I have had the misfortune of witnessing), and so I started developing this tool to automate this analysis.

Screenshots of the GUI windows and sample images are shown below.

BASIC WORKFLOW

• Installation and Set-up

  Download all files and folders from this GitHub page and place into desired directory. Then either:

    1) Set Path to this directory in Matlab
        Execute program by typing "QuantDRaCALA" in the console
        
    2) Use Matlab's Package App tool to package the program as an executable Matlab app  
        Start the program by clicking on the QuantDRaCALA app icon
  

• Pre-processing and Loading Images

  Read "Notes on Image Types" section below before using this program.

• Adjusting and Processing of Spots

  asdf

• Analyzing and Saving Data

  asdf

• Data Formats

  asdf

Main Functions

• QuantDRaCALA (.m and .fig)
• changeRadii (.m and .fig)
• splitFullPlate
• autoSegment
• ql2psl
• spotAnalyzer
• spotReIndex
• getSpotProps
• getSpotData
• singleSpot
• saveSpots

NOTES ON IMAGE TYPES (IMPORTANT!)

Sample images are stored as Quantum Level (QL) pixel scaling data, which is a log-transformed, non-linear pixel scaling method performed by standard phosphoimager instruments. The function ql2psl() [in ql2psl.m] performs a log-to-linear transformation on each pixel of an 8-bit or 16-bit grayscale 2D image to convert it to a Photo Stimulated Luminescence (PSL) value. The PSL value is the true measure of fluorescence from the radio-labelled ligand.

If images are already PSL-converted, then the QL2PSL algorithm will not run [function IN-PROGRESS] and data will be analyzed as is. Images in neither QL nor PSL format will not yield accurate data; however the trends in pixel intensity appear to correlate generally well with PSL-converted images, except in spots with low Fraction Bound values (FB < 0.05).

IN-PROGRESS FUNCTIONALITY

• addData_spotAnalyzer (fixing this functionality)
• deleteData_spotAnalyzer
• autoSegment (.fig)
• Data analysis and figure creation GUI (.m and .fig)

BUGS AND LIMITATIONS

Known Bugs:

• A seemingly random bug can sometimes inhibit the user's ability to resize the spots after the initial identification. Restarting the program and trying the analysis again should fix the issue, but the source of the bug is still unknown.

Limitations

• Draggable and Resizeable spots overload the image axis and make panning and zooming very slow. I haven't found a way to optimize dealing with the large number of objects. There is a strategy where spots are only visible if their coordinates are within the limits of the image in the main window axes, but I haven't implemented this feature yet.

Screenshot of the main QuantDRaCALA GUI window