LaCava Research Wiki

Materials and Notes

• tk180123_JLC_LD228_MT302_Hcheck_v2 - Checking %H-labeling and %H-mixing of cells expressing L1 from constructs MT302 and LD288, for use in I-DIRT screens. Our prelim look at this data encouraged us to adjust the amount of MT302 H-labeled powder in our mix (see below). We are waiting to finalize the calculation pipeline and determine if our response for MT302 was correct before also adjusting LD288.
• tk180302_JLC_MT302_24_12_Hmix_v3 - There are three different files in here: 12 reactions in duplicate as part of a 24-well screen. 12 single reactions done manually. 1 re-analysis of the mixing of the powders; the batch actually used in the 24 and 12 reaction experiments.

Order of events:

• determine in a single MQ run (a) %H-labeling of MT302 and LD288, (b) %H-mixing of MT302 and LD288 (v2), and (c) %H-mixing of MT302 (v3)

mqpar used for labeling and mixing v1

from %H-labeling we will determine the expected %H-mixing; how to the values from %H-mixing_v2 compare to expectation; how has from %H-mixing_v3 changed things? Which is more correct? From this we should establish a pipeline to assess %H-labeling and %H-mixing as well as providing a guidlinee on initial mixing proportions (presumably by mass) based on %H-labeling and protein mass/mg cell powder. We should also understand how to adjust a mixture proportions (presumably by mass) based on %H-labeling and protein mass/mg cell powder if an initial mix is not correct. Based on this understanding, we can address is adjustment of the LD288 powder is needed, make needed adjustments, and do the experiments. However, in some cases it may not be wise to adjust the powder, and may be preferable to instead compensate imprecise mixing in software by normalization (and below) - if the v3 mix is off, this will be needed.

• determine in a single MQ run (a) the results for the 24-well screen and (b) the 12 manual reactions. The results should be normalized based on the thouroughness of the %H-labeling and precision of the %H-mixing.

mqpar used for I-DIRT v1

this should allow (c) a rigorous statistical assessment of the I-DIRT performance in both screening and manual contexts and cross-comparisons of both; sufficiently compatible data (screen and manual) could also be combined, ultimately, if it improves signal detection. Apply the same to LD401 pipeline, however, as the mix cannot be adjusted (data is already collected) it may be necessary to apply normalization.