![]() Here we reduced this complexity by focusing on cellular organization-a key readout and driver of cell behaviour 3,4 -at the level of major cellular structures that represent distinct organelles and functional machines, and generated the WTC-11 hiPSC Single-Cell Image Dataset v1, which contains more than 200,000 live cells in 3D, spanning 25 key cellular structures. Understanding how a subset of expressed genes dictates cellular phenotype is a considerable challenge owing to the large numbers of molecules involved, their combinatorics and the plethora of cellular behaviours that they determine 1,2. Our software and hardware designs are fully open-source and include step-by-step build documentation to contribute to a growing open ecosystem of tools for high-throughput cell biology. Moreover, our strategy for automating the operation of a commercial instrument control software in the absence of an Application Program Interface (API) exemplifies a universal solution for other instruments that lack an API. Automation eliminates operator errors, standardizes gating conditions by eliminating operator-to-operator variations, and reduces hands-on labor by 93%. Our platform is built around a commercial instrument and integrates the handling and transfer of samples to and from the instrument, autonomous control of the instrument's software, and the algorithmic generation of sorting gates, resulting in walkaway functionality. ![]() Sorting large numbers of samples is laborious, and, to date, no automated system exists to sequentially manage FACS samples, likely owing to the need to tailor sorting conditions ("gating") to each individual sample. Here, we describe the development of an integrated software and hardware platform to automate Fluorescence-Activated Cell Sorting (FACS), a central step for the selection of cells displaying desired molecular attributes. To capitalize on this opportunity, new methods are needed to accelerate the different steps required to manufacture and handle engineered cells. Recent advances in gene editing are enabling the engineering of cells with an unprecedented level of scale. Other well level aggregation metrics such as average number (L) Validation with labeled data of the automated plate scoring pipeline. (K) Passed well filtering and ranking of total number of imageable wells per plate. Colony segmentation overlay output image with poor morphology regions outlined in red. (J) Steps of the CellProfiler colony segmentation and measurement pipeline. (I) Steps of the Ilastik pixel classification pipeline. (H) An example of the CellProfiler image inputs: bright field and the rough/smooth pixel probability channels. (G) An example of the smooth and rough pixel probability output channels from the Ilastik pipeline. (F) An example of segmented well image from a 96-well plate. (E) Boxplot of confluency of 6-well and 96-well plates seeded based on cell-count (orange) or on estimate from algorithm (blue). (D) Cell count numbers obtained from the cell counter were plotted against well confluency and shows positive correlation between the variables. ![]() (C) Formula used to calculate the well confluency. Scale bar is 500 m (B) Images acquired from a Celigo imaging cytometer were processed to segment and identify colonies in the wells and calculate colony features. Automated image processing to calculate well confluency, cell counts, and imaging well selection: (A) Brief schematic of 6-well plate with segmentation tiles (unstitched) outlining the colony area.
0 Comments
Leave a Reply. |