The numbers of cells in each field of view is therefore severely restricted. Since stage speed and the time for acquiring each image or stack limit the number of fields that can be followed, the increase in cell density in ALCATRAS results in a direct increase in the total number of cells that may be recorded. For this reason, we have been able to raise the scale of data collection by an order of magnitude. To gather large datasets, a device must also be loaded efficiently. Our devices can be fully loaded using flow conditions similar or identical to those used during data collection. Alternative devices are either loaded at low efficiency or require special conditions such as high pressure or device centrifugation, which have the potential to affect cell behaviour. Our device opens up new possibilities for biological research. The NSC 663284 ability to retain mother cells while measuring responses to environmental change will allow studies of the physiology of ageing. Precise experimental control of the cellular environment will allow interrogation of the mechanisms by which cells process signals, and so provide insight into cellular decision-making and determination of cell fate. In systems biology, the ability to obtain time-series of thousands of cells will improve model fitting and selection. Such large datasets, coupled with environmental control, may also allow studies of rare events, such as the emergence of drug resistance. The focus of cell biology research is shifting from the study of populations to the study of individuals. Our microfluidic device and others like it have the potential to make single-cell studies widely available and statistically robust. Image segmentation, cell tracking and data extraction was performed using customized code in Impentamine dihydrobromide Matlab. Traps were tracked through time using an approach based on normalized cross-correlation. Individual cells were identified with a trained classifier using a support vector machine. Following identification of a cell, the outline was detected using a circular Hough transform. For both cell cycle and glucose limitation experiments the nuclear GFP localization was determined by the protocol of Cai et al., which is unaffected by photobleaching and changes in numbers of fluorophores. For the cell cycle experiments, we applied a threshold to the rise in nuclear localization between successive time points to identify nuclear entry of Whi5p-GFP. For the glucose limitation experiments, we calculated the threshold for determining whether a cell had nuclear localized Msn2p-GFP using the mean and standard deviation of all cells for each experiment.