Monday 12/30/13

9:20a -7:00p

Goals

attempted software update to do bead-averaging on-the-fly. Failed.
STORM2 software several updates behind, frame objects have been substantially redefined.
More consultation with Hazen, updates will require some substantial reconfiguring. Needs to wait until Colenso’s data collection is finished.
github not yet set up on STORM2, let’s just try this manually for now.

Starting with image _0_0, 647 channel having trouble being well resolved in STORM images. Not sure why?
- observation 1, 647 hardly visible in conventional image. I know this slide had good labeling, clear conventional dots. Not sure what’s up.
trimming down the number of frames seems to improve the quality in some cases. I think some images exhausted the low level labeling in the first three or five thousand frames and spend the rest of the movie accumulating background
background is legitimately quite high in theses movies.

working on building a demo-data set
don’t really have impressive multi-color STORM images that I’m ready to share openly with the public.
building a demo data from down-sampled confocal data of Drosophila embyro — this should illustrate most of the points.

Continue reading Kieffer-Kwon et al Cell 2013.
still, a very nice data set. Best display of ChIA-PET data I’ve seen, the chip-seq tracks with the red lines linking peaks works well. The promoter enhancer culstering graphs and the complete array organized by complexity are also very nicely done.
Types of PolII groups — extragenic – tethering promoters to distal enhancers; intragenic, connecting promoters to gene bodies — (don’t you think these are intronic enhancers? they are rather clear peaks from intronic sequence more than general spread across the gene body) + promoter-promoter and enhancer + enhancer.
Though on closer examination I see that they see a clear, though subtle increase throughout the gene body as well of transcribed genes, so maybe this is what is referred to. In which case I think it is very likely enzyme accessibility.
how much of this is open chromatin driven? Maybe open chromatin is the biological basis of selection?
okay, also later on they do refer examples of an intronic enhancers. So you could call it 5 classes and be clearer, or not define enhancer contacts as necessarily ‘extragenic’.
90% of enhancers linked to a single promoter, less than 2% linked to more than 2, max = 7. (what do you expect by chance as a function of local density of open chromatin regions?)
I wonder how many of these promoter promoter contacts are promoter proximal enhancer elements?