Recall that the Flow Cytometer "sucks" cells from liquid sample and pass them through a flow chamber where they hit a laser beam which allow to measure both fluorescent signal and the size of the cell. The HyperCyt loader is a machine that allow to collect samples from a multi-well plate and send them to the FlowCytometer. You can see a movie in the original blog entry. Thus, the FlowCytometer receives a sample that consists of 1.5 seconds of a "sip" into a well, followed from a 1 second sip from air. At the end of a 12-well row, there is another step of sipping water to clean the pipes.
Today Ariel (our Ariel, not to be confused with Ariel from Merkel who did the Flow Cytometer upgrade), performed a test to see the whole thing in action. Now that we have upgraded the Flow Cytometer acquisition hardware/software we can start examining the results. To see how good they are, we plot the time a cell was collected vs. the measured value. In situations where there were no cells we will not see points.
When Ariel ploted the time vs. Forward Scatter (a measure of the extent to which the laser beam was scattered by the cell, which is a proxy of cell size), we see a band for each row of wells.
Zooming in to look at a particular band, we can see that it consists of intervals that correspond to a sample separated by blank intervals (air bubbles).
We can observe several things here.
First although each "sip" is of 1.5 second, some of them are spread out on longer measurement time. This is probably due to the fact that the bubbles are spread out by the transition in the tube, and the first bubble might have more space to spread out.
Second, although air bubbles are easily detectable by the lack of any events (cells), the empty wells Ariel added (marked by Medium and water) do contain some cells. These are probably carry over cells that got temporarily stuck on the tubing and then released by the flow. We can see that by after the wash by one well, the subsequent water well has much fewer events. This will give us a sense of how much carry over we can expect.
If we look at the GFP measurement for these cells we see similar bands. Here Ariel used strains of different GFP intensities, and we can see from the plot that they differ.
For example, the "no-gfp" strain shows very low level of GFP (basically auto-florescence), while some of the other strains (e.g., Sod1, Rpl25A) show very high levels.
This is a very first run out of many we are going to perform. Clearly there are issues to deal with, such as how to automatically parse the output into wells, how to reduce the amount of carryover, how reproducible are the results and so on. However, I am very excited to start seeing all these different things start to work together.