sonyahanson
commented
8 years ago Lucelenie had asked me a few days ago how the mg/L expected scale-up conversion was made, and I couldn't answer, so thanks for clarifying this!
So, just to summarize:
- You would like us to fix the calculation to use 0.9 mL of culture instead of 1 ml and change to using this mg/L value in the table and figure legend for the kinase tree.
- You would like us to change the expression cut-off we use to 2 mg/L, according to this new calculation, and change any current occurrence of the 68 expressable kinases to whatever this new number ends up being.
jchodera
danielparton
It looks like the current manuscript has some major confusion over the numbers we report for expression yields.
The Caliper GX II yields are reported by the MacroLab in "ng/ul", which actually represents the number of nanograms of protein per microliter of sample loaded onto the Caliper. The sample comes from the eluted protein that came off the beads in the Nickel Buffer A incubation and wash (120 uL). Presumably, a few uL (how many) of this were loaded into the Caliper for quantification, and the final reported number is the concentration of this 120 uL eluate. The "ng/ul" number is not informative without reporting how many total uL of eluate we got from the culture volume, and how much that culture volume was.
The original culture volume was 900 uL, so Danny and I had been converting the ng/ul number to "ug/mL of bacterial culture" by multiplying by (eluate volume) / (culture volume). This is why the online table lists number 12 M3K5_HUMAN_D0 as 105 ng/ul but 12.6 mg/L expected scale-up: we appear to have calculated this as (105 ng/ul eulate) * (120 uL eluate) / (1000 uL culture) = 12.6 mg/mL culture (where you see we have erroneously assumed we had 1 mL culture instead of 0.9 mL culture). There seem to be some other issues with rounding in the online table from Danny as well----for ABL1_HUMAN_D0, the mg/L number 2.26 is truncated to 2.2 instead of being rounded to 2.3.
Can you guys fix the manuscript with this in mind? I'll check in my edits in just a few minutes, and then you can have a go at it. At minimum, we need to make sure we're talking about the right thing when we quote the yield or how many kinases expressed with a certain yield. Even better would be to draw a threshold of, say, 1 or 2 mg/L expected scale-up, since it isn't very reasonable to use this method for the kinases expressing way down at 0.27 mg/L (2 ng/ul eluate). That will drop the number of kinases we report as expressing, but will make it less embarrassing than suggesting people should use the kinase constructs that barely express.
Tagging @Lucelenie, @sonyahanson, @steven-albanese