In a previous post I showed how to use Omni plates to titer bacteria. Another useful application of OMNI plates is to plate out transformation mixes. In a high-throughput (96 well) cloning process it is easy to run In-Fusion reactions and transform bacteria in 96 well format. Vendors provide competent cells for chemical transformation in 96 well format. The bottleneck is plating and picking colonies. OMNI plates can be used to plate out 96 individual cloning reactions per OMNI plate.
The bottleneck then shifts to the next activity - picking colonies for sequencing. A Qpix2 colony picker can be used for picking. The Qpix deck will hold 2 OMNI plates using an Omni tray holder:
The software identifies a single region per plate (two regions total) and randomly picks a user defined number of colonies from each region. Ideally one would want to define 96 regions per plate so that the number of colonies per region can be specified - this is not allowed. However, each colonies location (X|Y deck coordinate) is recorded in an XML log file, and with a bit of hacking, the well location can be assigned to each colony.
The first step is to determine the how to interpret the coordinate system of the deck. I start by placing an agar filled OMNI plate on the deck of a BiomekNX liquid handling robot. Load the head with P20 tips and impale the surface of the agar.
The indentations on the surface of the agar are (mis)interpreted by the software as a colony and the X/Y cordinate will be reported. From this data one can determine the X|Y layout of the deck.
Now I can determine the coordinate boundaries for each of the 96 regions of interest:
More generally I can determine the coordinate layour of the entire deck. The coordinate system appears to be upside down, not surprising given that well A01 for each plate is bottom right corner. Apparently the engineers were standing at the back of the instrument during the design process. This has not been corrected in the newer version.
Now knowing the coordinates of the center of each well, I can assign an area to each well, and a well id to each colony that falls within that area.
I will assign an anchor coordinate and calculate everything as an offset to that. What if a PM or service call changes the deck slighly? I would like a way to calibrate. I find that I can use an ABI PCR plate holder as a calibration tool. Place the holder in the OMNI plate, scan, and the image can be used to assign an anchor coordinate. Because row H is too close to the edge of the image, “colonies” (actually the holes in the ABI plate) are not identified, so I use well G01 as my calibrator.
I can then calculate the offset between the ABI plate and the impaled OMNI plate. Should the coordinates of the ABI plate well G01 ever change, the change can be incorporated as an offset in the scripts.
Using this method I can assign source wells to the colonies I have picked and look at the pattern of deposit into the destination plate.
In this image the X|Y coordinates correspond to the OMNI plate e.g. OMNI plate well A01 is in the upper left corner, and the well ID is from the source plate. So the transformation mix from well D12 was deposited at position A01 on the OMNI plate. Looking at the pattern, I see no logic to it. Molecular Devices claims the picking order is optimized for speed.