|
Excerpts from the upcoming 2007 report “Image-based Assays in Drug Discovery: The Next Five Years”
Current issues facing image-based assays: Cell variability—various advances in stem cell technology promise to revolutionize the ability of drug companies to screen against relevant target populations of cells. However, variability in the handling of cells often leads to variable results. Automation of cell handling to the greatest extent possible is the key to obtaining good performance in a screening context. Advances such as cryopreservation of synchronized cell lines pioneered by CMT have further reduced variability.
Instrument and image processing expertise—despite their age within the scientific enterprise, microscopes are precision instruments that many operators are fundamentally unfamiliar with. Their proper operation and the accuracy of data collected requires that users be trained in operation and calibration procedures. Quantitative intensity and color measurements require precise calibration with industry standards such as fluorescent plate standards or beads (flat field, wavelength, and lamp stabilization and output measures should be tracked routinely). Frequently, vendors are simply unable to provide users with the background necessary to properly calibrate and use these instruments. Additionally, while use of pre-configured image processing algorithms is common, it is impossible to design one image processing routine that will accurately analyze the entire range of biological variability that is encountered when there is a shift in cell types or conditions. The ability to do basic customization of image processing routines is key to result accuracy. Users should also be familiar enough with the basic concepts of image processing to be able to assess the performance of pre-configured algorithms.
Plate readers vs. microscopes—plate readers lack the image processing and computational overhead of imaging systems, and usually report a single parameter quite rapidly. Care must be taken to determine if the added information associated with imaging assays is worth the extra time and effort. However, it is worth noting that, depending on the instrument and the magnification used, the throughput speed achieved by many imaging instruments can be equivalent on a per plate basis to plate readers.
Consolidation assays are underutilized—consolidation assays simply combine pre-existing assays into a single assay, and are not novel in themselves. Yet by capitalizing on the information already on the plate they can provide a significant advantage. By shifting surrogate endpoints normally measured with custom reagents to morphological readouts (such as cell shape as a readout of cell health) or quantitative dye measurements (DNA binding dye intensity as a measure of cell cycle) they can lower the per well cost of the assay and possibly eliminate other assays. Typically the recovered cost, time and color channels can then be used for more expensive reagents such as antibodies. However, this type of assay is currently underutilized in industry.
Interpretation of complex data—cells in vitro are complex organisms, and often the complex data they generate can be overwhelming. For example, investigators have difficulty interpreting differences in response between different cell types. Patterns of response across different cell lines is another form of “meta-data”. The interpretation of this meta-data is benefited by increasing amounts of information available in public databases.
|