Inspiration
We wanted to measure communicating synthetic consortia’s fluorescence over long periods of time to best characterize their interactions and their responses to external signals (i.e. autoinducers). If we were to do it manually by flow cytometry or plate reader experiments, it would require too much manual labor. To get the kind of longitudinal data we wanted, we needed to design a software that could compile hundreds of microscopic images and analyze fluorescence on a single-cell scale. With thousands of images to analyze, we quickly realized that the experiment file sizes could easily exceed 100 gigabytes, making it crucial to develop a fast and efficient software solution capable of processing such large datasets without relying on superclusters.
ActivityPlot
The ActivityPlot is simple yet a powerful tool designed to process and analyze thousands of microscope images. Our designed tool cleverly uses the Sobel algorithm to find the cell- filled mono layer chamber of a PDMS chip in a given microscopy image. This software is user-friendly and supports common file formats such as ND2 files for input and CSV files for output, making it accessible to researchers and scientists in various fields. The software is deliberately designed to run efficiently on standard laptops and desktop computers, eliminating the need for high-end supercomputers or specialized hardware. This ensures accessibility to a wider user base with modest computing resources.
The flow chart above shows that the steps ActivityPlot takes in order to process the Mircoscopy data.
Workflows
- User selects ND2 files or defines a directory containing multiple ND2 files.
- User defines the output directory for CSV files.
- User defines title, normalization RPU values.
- User initiates the analysis process.
- The software processes each image
- Results are saved in CSV files and outputs the plot in the specified output directory.
Figure 1: Running ActivityPlot on a IGEM practice experiment using the terminal
As depicted in Figure 1, an experiment spanning a duration of 67 hours and involving 13 mono-layer chambers generates an extensive dataset comprising over 10,000 images. The conventional approach of manual data processing methods would have entailed days' worth of laborious efforts. However, with the deployment of the ActivityPlot tool, the entire experiment is efficiently processed in a mere 15 minutes. This substantial reduction in processing time is a testament to the software's exceptional computational efficiency and its ability to streamline the analysis of voluminous image datasets.
Key Features
Batch Processing
In the context of extended-duration experiments, it is common to encounter a multitude of data files, each representing distinct segments of the experiment. The software automatically processes these files and effortlessly combines the data into a single, coherent dataset. This automated data integration eliminates the need for manual effort and ensures a unified dataset, making it a valuable tool for analyzing complex, extended experiments.
Easy to use
To make Activity Plot accessible even to those less familiar with computers, we've taken several user-friendly measures. Our scripts are well-documented and have README files for clear guidance. They are exclusively written in the high-level Python language and rely on commonly used nd2 and CSV formats for input and output, ensuring simplicity. Moreover, we are currently developing a graphical user interface (GUI) to make the software even more user-friendly, especially for individuals who may not be well-versed in command-line processes.
Future Direction
In our future plans, we envision expanding the capabilities of the software to be adaptable to various microscope data file formats, ensuring flexibility for researchers using different microscopes. Moreover, we aim to develop a web-based version of the tool, enhancing accessibility for scientists worldwide who are interested in analyzing cell fluorescence within microfluidics. Our vision is to create a versatile and globally accessible resource for the scientific community, simplifying the process of studying cell behavior and interactions in microfluidic environments.