A computer system that automatically tracks the movements of proteins within a living cell has been developed by a team of biologists and computer vision experts. It could save researchers the hours often spent analysing microscope images by hand, to determine the way a cell works.
The system, called CellTracker, automatically analyses a series of still digital images captured through a microscope.
Doug Kell at Manchester University in UK, the lead biologist involved with the project, believes the system could dramatically speed up studies of cells鈥� function. 鈥淢ost people just fix cells [in one place], which kills their metabolism,鈥� he told 快猫短视频. 鈥淏eing able to look at live cells over time is much more helpful.鈥�
Advertisement
The system uses image recognition algorithms to identify the membrane marking the edge of a cell as well as the one enclosing the nucleus, which contains the cell鈥檚 DNA. It can track the movements of these features in multiple cells simultaneously.
Fluorescent dyes
CellTracker can also monitor different proteins contained within the cell, providing these have been labelled with different fluorescent dyes, and produce graphs showing fluctuating levels of these proteins over time. This could help biologists understand the way cellular proteins affect cell function.
This video shows three cells and their nuclei being tracked by the software (AVI file, 10MB). The green colour seen inside the cells is a labelled protein.
Analysing cell proteins is normally a time-consuming task, Kell says: 鈥淧eople have to cut around the cells in each image, which is torture. We estimate CellTracker can do in half an hour what would have taken 12 hours before.鈥�
The software has been publicly released for other researchers to use. Kell鈥檚 team plans to use it to track cellular signalling proteins. These communicate messages between different parts of a cell and are crucial to cellular function.
Manual work
Jon Lane, a biochemist specialising in microscopy of live cells at Bristol University, UK, says the computer system sounds potentially useful. 鈥淲e do a lot of manual work here, and it鈥檚 really very tedious,鈥� he says. 鈥淲e would definitely think about trying this out.鈥�
Other software tools can help biologists record cell movements, but Lane believes CellTracker offers something new. 鈥淭his software鈥檚 ability to track cells and find their edges even when they move about or touch is its strength,鈥� he says. 鈥淚鈥檝e spoken to some colleagues that have used it, and they鈥檙e very impressed.鈥�
However, Lane doubts it will be accurate enough to use for very detailed cellular analysis: 鈥淭o look at very small movements on the scale of microns, you鈥檇 still have to do it almost by hand.鈥�
