User:Steven J. Koch/Notebook/Kochlab/2010/05/04/Planning for data automation

From OpenWetWare

< User:Steven J. Koch | Notebook | Kochlab | 2010 | 05 | 04
Revision as of 00:09, 5 May 2010 by Steven J. Koch (Talk | contribs)
(diff) ←Older revision | Current revision (diff) | Newer revision→ (diff)
Jump to: navigation, search

Steve Koch 00:06, 5 May 2010 (EDT): Today, Andy, Larry, and I made some plans.

  • Larry fixed up automated tracking so that the last bad point is not included.
  • He also thinks he improved it so that MTs are not tracked multiple times.
  • We have an idea that we can remove duplicate tracks very easily via some kind of comparison algorithm (similar to SDM).
  • With all of the above, and the robustness of the kernel density estimate method (not sensitive to huge stray points, like mean would be), we all feel that it may be possible to fully-automate the data analysis. This will have benefits
    • Save a bunch of time before conference
    • Allowing us to get more data, and compare two methods:
      • PDF of all concatenated data
      • PDF of (all) peaks from individual data sets.

I'm actually optimistic that both methods will work. Larry pointed out that differences between the two may tell us something. I'll need a day or two (hopefully that's all) to write some software to do that, while Larry is going to switch back to the kinetic modeling software and tweak it a bit so we can run some simulations before the conference that may help in interpreting the experiments. At the very least, he'll be able to re-do some of the interesting ones he did and be solid for the poster. I think mostly what we'll show is that we have a very solid experimental and computational platform for studying kinesin. Considering that at this point last year remodeling our lab hadn't begun yet and we didn't have kinesin until August, that's a huge success. But that's not all we're going to show, since I think we can tell some science stories. It's just going to be tough to make them complete stories:

  • Andy has data to make recommendations for passivation. This builds on Velma et al. work, and I thin is technically interesting for people using the assay for science or devices.
  • Heavy-hydrogen and heavy-oxygen (especially the latter) water may actually be a good knob for studying the system. Heavy-hydrogen may have beneficial effects for devices.
  • The kinetic simulation work points out inconsistencies in the literature (such as rate of Pi release). This would be good to hash out with the experts.
  • Betaine (and sucrose) have dramatic effects on the rate of transport. Also perhaps on processivity (though that is tough to see in this assay). We're not sure, though, based on some conflicting literature whether it may be a viscosity effect. I'm worried, because sucrose (much more viscous than betaine) had a bigger effect, at least the first time Andy tried it. Compared with heavy-oxygen water, I am less confident. Which is ironic, because I've flip-flopped twice on this. The heavy-oxygen results seem like potentially the simplest to interpret, so I'm very glad Andy did those experiments so quickly!

I'm extraordinarily nervous about being around all these kinesin and molecular motor bad-asses in two weeks! http://cnls.lanl.gov/molmot/agenda.html

A teeny bit of analysis

  • In the process of getting tracking working on the new computer, I was able to to a teeny bit of analysis on Andy's beta-casein data from March. But I ran out of time due to tee ball game. Sam got a great out at second. My analysis from today can be found here: http://kochlab.org/files/data/Microtubule%20Tracking%20Data/2010%2005%2004
    • The most notable thing is that there is a very quick microtubule. I think it is real, because I saw it in another that I wasn't able to track because it was too small.
Personal tools