AMEBaS: Automatic Midline Extraction and Background Subtraction of Ratiometric Fluorescence Time-Lapses of Polarized Single Cells

This method quantifies the spatiotemporal dynamics of polarized cells along its midline expressing fluorescent reporter or dye for specific entity of interest such as actin, ions, or vesicle dynamics. The version of automation eliminates biases and permits scalability, especially for tasks that are digitally done manually and can become time-consuming and subjective. To begin, open Jupyter Notebook and read the time-lapse files, either by accessing the Google CoLab interactive notebook homepage, or downloading and opening the AMEBaS_local IP YNB from GitHub.

To prepare the directory setup for the input and output data using the local version, place the fluorescence time-lapse as a TIF or DV file inside a folder named data located in the program's root folder. Create an out folder to receive the generated data, then run the setup code block. If using the notebook on Google CoLab, execute the setup code block to automatically generate data and out folders, then run the file input code block to read the time-lapse data by clicking the play button.

If utilizing Google CoLab, upload the time-lapse file to the data folder by clicking the choose file button. Next, choose to generate additional outputs for each step by setting the verbose parameter to true or false. To detect the main cell and segment from the background, run the single cell segmentation code block by clicking the play button to automatically separate the cell of interest from the background.

Adjust the sigma value in the sigma variable to fine tune the smoothness of the segmentation mask. The default value is 2.0. Now set the variable estimate to false for storing the threshold estimated directly from the ISO data, or true for smoothing it across neighboring frames using local polynomial regression.

Adjust the n_points variable to fine tune the function with the default value being 40. To trace the midline along the cell extension, run the cell midline tracing code block by clicking the play button to skeletonize the cell using Lee's method and extend the tip of the last skeleton through linear extrapolation. Next, select the tracing option for the midline by adjusting the complete_skeletonization argument to trace either on the last frame or once per frame.

Adjust the interpolation_fraction variable to set the fraction of points in the skeleton for interpolation during extrapolation. The default value here is 0.25. Next, modify the extrapolation_length variable to determine the length of the midline extrapolation.

The default value is minus one, which extends the skeleton to the nearest edge. Now run the first data visualization code block by clicking the play button to automatically generate kymographs for both channels. Choose the Gaussian kernel size for smoothing by adjusting the kymograph_kernel variable.

To display the intensities of nonextended skeletons properly, a custom color map is needed for their capped kymographs. Adjust the shift_fraction variable to choose the fraction percentage of intensities that will be designated as the background color, black. Then run the second data visualization code block by clicking the play button to generate a ratio metric kymograph and time-lapse automatically.

Adjust the switch_ratio variable to change the order of channels used as the numerator and denominator during ratio calculations with the default being false. Check if the ratio time lapse requires additional smoothing with a median filter pass by adjusting the smooth_ratio variable. By default, this is set to false.

Next, choose to either remove or keep outliers resulting from low signal in the denominator channel by adjusting the reject_outliers variable. Outliers are marked as values 1.5 times the interquartile range above the third quartile. Finally, choose if the background and the ratio metric output needs to be exported by adjusting the variable background_ratio.

The default is false to replace the background with zeros. AMEBaS tested on datasets like pollen tubes expressing the calcium reporter cameleon, pollen tubes expressing the pH indicator florin, root hairs expressing the calcium reporter NESYC 3.6 worked successfully despite differences in the growth direction, imaging techniques, fluorescent reporters, and cell types. Remember to execute the set of block so you can install the required packages.

Additionally, it's crucial to adjust the sigma values specifically for your images so you can get more accurate results. The resulting kymographs then can be used to analyze the spatiotemporal dynamics of the solve interest. Additionally, you can calculate it finding method, growth and time series to analyze the right solution.