Alistair's Lab Notebook

9:30 am – 8:15 pm, 9:30 pm – 11:30 pm

STORM

• continued imaging ANT-C G6 in WT cells (imaging finished at 6:17 pm)

Staining

• rinsed new antibody stains (CycA and Ph)
• CycA (after formamide) now only stains the cytoplasmic fraction
• 488 is extremely bright
• still distinguishable by confocal (not so obvious by eye-pieces).
  
• This Ph still is not nuclear in these Kc cells
  

Ph troubleshooting

• added 60 uL blocking buffer to dried out / empty tube of old Ph antibody. Maybe there’s still some left. Used 20 uL of the resuspended anti-serum to stain WT cells.
• incubated 1 hour at room temp and then overnight at 4C

STORM

• CycA stain is essentially undiminished from this morning. Still very bright. C03 stained clearly as well
• BXC stained well but the replacement F06 I used did not stain well
• reverting to the previous sample (here BXC is a little below what I desire and in 750, though F06 is bright).
• tomorrow I should try staining BXC-P1-647 + F06-P3-750.

Fly husbandry

• flipped stocks.

Presentations

• working on lab-meeting presentation for Friday

Communications

• send Hi-C summary to XZ
• reply to CW about cell cycle
• reply to NF about paper

11ish – 4:30 pm, 8:00 pm – 11:00 pm

remotely 5:30-8:00

Analyzing Hi-C data

• effect of flank size on isolation score:
  
• scaling exponents within a domain have a troubling dependence on domain length.
  
• drafting description of Hi-C comparison
• see Revision sub-folder in Chromatin/Figs for latest version

9:00 am – 10:40 pm

Live imaging

• live imaging of telomere-mmaple3 construct on STORM2
• need to correct live drift so we’re stuck with auto-correlation for these, no need to acquire beads
• blinking pretty well in current settings
  • 800 mW set point on 561, used at 40% power on AOTF
  • 2% constant 405 after first ~2000 frames to let the background bleach down a bit seems to give good blinking
  • slowly ramp to 3% 405 by the end of the movie (60,000 frames)
  • flipped out to 2D (conventional spots look better)
• autocopying to STORM2
• cleared disk space on STORM2 C
• Bogdan will take the analysis of these from here

Fixed imaging

• imaging ANTC-G6 in WT cells
• settings
  • flipped back to 3D
  • MEA buffer (5uL) no COT, fresh MEA in MEA dissolving buffer, fresh Glox.
• staining looks a bit dim/low contrast, maybe should have washed longer.
• temp might have been a bit low as well (75 instead of 78 — the glass actually cools down the block).
• Blinking looks good, hopefully still enough localizations for a robust characterization of domain size and overlap.

New stains

• anti-CycA on formamide treated WT cells from 8/11 fix.
• anti-Ph on WT and Ph KD cells from 8/11 formamide treated
• primary antibody 1.5 hours

Hi-C analysis

• working out intermixing model and illustration
• see Chromatin/Figs/Revision
• isolation of blue domains from flanking sequence for all domains not too robust

9:30 am – 7:00 pm,

Lab meeting prep

• working on slides for Group B presentation, 9:30 am – 1:30 pm

Two-color data analysis

• interesting results from ANT-C G6 quantification
• median overlap ~25%, median entanglement ~0.06
• need to make figures still

Meetings

• requested to make MERFISH slides for grant meeting
• practice talk to be given Sept 4 (next Friday)

Sequencing prep

• ran tape-station on 1:100 dilution samples
• looks good, library length peaks are 297, 302 and 307.
• Problem except I cant’ connect to the data folder: https://data.rc.fas.harvard.edu to get the raw data to post here.
• submitted request to RC help for troubleshooting.
• computing final concentration (now that we have average library length).
• 600 nM * (450/300) = 900 nM starting. Desired 10 nM. So my 1:100 dilutions are around 9 nM, pretty good.
• submitted sequencing request (still did not submit physical samples, should follow up with Christian).

To do

• reads per domain for all domains (by color) compared to green domains
• analyze the BX-C F6 data
• images of the BX-C F6 data for presentation
• quantification comparison of BX-C F6

For later

• restain Ph with shorter fix time
• try Ph stain on MeOH fixed cells

9:30 am – 9:00 pm

Goals

• ChromatinCropper2 two color data for ANT-C G6
• Spot finding for two color BX-C F6 data
• ChromatinCropper2 two color data
• tape station of sequencing prepped library for Ph-KD and control
• qPCR of sequencing prepped library for Ph-KD and control

STORM

• stop STORM run, clean up STORM2 for Yari
• analyzing data on STORM2
  • shouldn’t launch more than 30 jobs at once due to read/write limitations through the USB3.0 port on Morgan — makes interacting with the disk impossible.

Data Analysis

• troubleshooting issues with chromatic alignment
• fixed issues with chromatic alignment
  
• Cropper2 not applying warps properly: previous warpfile was still hardcoded, not being passed
• fixed warp for alignment of conventional images.
• still need to make sure warp gets applied to STORM images.

Deep seq prep

• rerunning Kappa quantification
• kappa standards 1:6 in triplicate (A1-6 through B1-6, C1-6) + No template controls A7-C7
• original dilution 1:100
• serial dilution (in triplicate)
• 1:100,000, 1:1,000,000, 1:10,000,000
• 10 in 990.
• just used median in fit. linFit(x) = p1*x + p2; p1 = -0.3005 (-0.3083, -0.2927), p2 = 4.989 (4.823, 5.156)
• looks good
  
• 1 in 10,000 dilution is ~60 pM. so original is 600 nM, assuming average molecule size of 452. Otherwise correct by multipling by 452/aveSize.

Genome analysis

• figured out header structure and wrote missing header for modENCODE Kc167 SAM data.
• saved as dm3_SAM_header.txt on Morgan on \\MorganData1\Chromatin\2015-08-25_KcENCODEdata

9:30 am – 10:15 pm

STORM

• finish overnight run
• switching still looks decent
• setup data to analyze on MORGAN.
• updated z-calibrations for both 647 and 750 parameters
• fitting bead data for drift correction
• all jobs launched before 1pm (probably need another 2 hours to finish running)

New experiments

• stained BX-C and F6 (using previous mixed sample)
• denatured at upstairs heat block at 78 (temp reasonably sensitive to adding water, use with care)
• hybridized from 10am – 7pm at 47C
• rinsed and imaging in MEA and freshly made Glox (left glox made yesterday out last night, tossed it)
• spots switching much better in 750 (Ben also fixed the TIRF)
• collecting ~55K frames 750 (30Hz, camera won’t do faster) and 65K frames 650 overnight

Ph-project

• assemble figures, send to NF
• latest complete version sent to all authors

Analysis of Green data

• mapping total reads per region
• Morgan connection got really slow

counting total reads per region

I used to use a “GenerateStrandCount” function you had written (I believe in matlab-functions).
I presume this has been replaced by “GenerateCoverageVector” or “GenerateSequenceDataStructure”?
Both of these try to find a “SequenceDictionary” field. The structure I get from BioIndexedFile on my SAM file doesn’t have this field.
I believe this is because my SAM file has no header. That pissed off cufflinks too, though I eventually got it to run. modENCODE doesn’t seem to believe in headers.

brute force approach

• using fseek to count reads, no fancy coverage vector for whole genome
• running over night

To do

• format figures into combined PDF for Ph-Polymerization paper
• Analyze Hi-C data
• Organize presentation of Green data
• Analyze two-color data

9:30 am – 5:15 pm, 8:20 pm – 11:30 pm

Ph-polymerization

Manuscript to do

1. STORM section, reply to notes in main text
2. Finish Layout Supp figure 2 and Supp figure 4
   • added supplement captions form both
3. Check the legend to Figure 5 (new model) and sup. Fig. 11 (rest of model fig)
4. Compile everything into a PDF
5. Check response to reviewers.

STORM

• imaging on STORM2
• 2 color
• MEA COT buffer
• sample quality (ANTC-750 + G6-647)
  • still an odd background nuclear glow in 647 channel (not related to STORM-dye switching, uniform, unbleaching. Still suspect some contaminate Draq5 or something of the sort. Maybe the formamide overnight treatmetn helped drop this down, the STORM switching on top of this background looks pretty good.
• 561 beads just switch dark / bleach quickly in the MEA COT buffer
  • swapped fresh buffer, added fresh beads. Conventional imaging beads nice and bright
  • start STORM imaging, and beads go dark pretty quick and don’t come back
  • I recall this happening before, and only lasting the first few movies
  • indeed the beads do come back bright, but now the 750 switching looks bad.
  • probably the buffer quality is decreasing, favoring the beads and disfavoring the dyes again. Unfortunately the switching quality in 750 is a no-go at this point.
  • I want to try this again in BME
• switched buffer to BME.
  • beads don’t recover very well. They aren’t as completely dim but its still pretty bad. Should replace with fresh beads again.
  • 750 switching not as bright, and most dramatically, not as switchy — fewer blinking dyes for way less long. Need to recalibrate all my ramp files
  • using substantial 405 I can get some switching. Let’s see how this goes
  • next I’ll try MEA alone, no COT.
• switched to MEA no COT
  • first movie, 750 dim but blinking, 45K frames instead of 70 (better than the BME, but we’ll see if it lasts)
  • at least the newly added beads stay bright
  • running the MEA overnight.
  • still a total unusual amount of background in the 647 channel

Cell culture

• Psc null cells growing well. I have 3 vials in freezer so hopefully those are fine if I need them again.
• Terminated Psc null flasks
• trimmed down to 1 KC in SFX flask and 1 KC in Schneider’s flask.
• interestingly the Schneider’s cells are now adhering better than the SFX

Seq Prep

• Tape station training
• tape needs to be reserved online
• ordered 1 tape for next set of experiments
• HS DNA is the recommended tape for library prep.
• protocols sent in email.

9:30 am – 7:00 pm

Ph project

• rereading and revising manuscript
• sent comments back to team
• revising layout of Fig 5.
• revising response to reviewers
• sent updates back to team

STORM imaging

• Jeff realigned lasers on STORM4 table, fixing the 90% power loss on the 647 laser due to poor fiber coupling.
• quadview insert and master need to match. Swapping these may have damaged the mirror on the quadview for the 560 channel on STORM4 (run with STORM2’s insert) and mirror for the quadview on STORM3 (running with STORM3’s insert).
• can probably run with shutters…

Chromatin analysis

• ChromatinCropper is crashing at the scatter command for large fields of view
• fixed, change to plot
• analyzing Green domains
• forgot to change npp back to 160, will do in post analysis

9:30 am – 1:30 am

Microscope troubleshooting

• might be able to try 2-color imaging in STORM5 with George E and shutters.
• doesn’t work. No emission filter to block 405
• 2 color conventional images of my stains look great though.
• trying to image on STORM4
• issues with quadview – 561 channel not working

<

del datetime=”2015-08-21T21:53:51+00:00″>

Writing

• revise draft of Ph-manuscript and send replies to Nicole and Ajaz

RNAi experiments

• Day 4 for KD. Lyse cells, perform RNA extraction.
  • collected 7 tubes of cells, concentrated in spins
  • just froze cells in lyses buffer at -80C. Will continue these when I have time.
• fix some cells for imaging backup
  • prep cover slips
  • plate cells, allow 3 hours to attach
  • fixed 4 coverslips of Ph KD and 4 coverslips of mock.
  • these may be necessary for continuing the multicolor experiments

Meetings

• discussion with XZ about grant, 11am
• update on revision experiments, 1pm

Live imaging

• TIRF2-mmaple on STORM4
  • sample don’t look promising
  • snapped some images, saved on STORM4, need to copy over for update meeting
• TIRF2-GFP on STORM3 (Bogdan, 6pm)
  • samples not well adhered (polylysine treatment was omitted).