Preliminary data

10 views
Skip to first unread message

Ross, David J. (Fed)

unread,
Dec 9, 2016, 2:13:59 PM12/9/16
to sbsc-flow...@googlegroups.com

Here’s some example data I got off the Attune yesterday at different flow rates.

Data is for the mCherry only E. coli strain that Nick sent.

Growth protocol was shorted for initial test: grow overnight, dilute 300x, grow 3 hours, put on ice, dilute 20x for measurement.

 

For each flow rate, I’ve attached plots of:

(left) the SSC vs. FSC as a density plot;

(middle) the SSC vs. FSC as a dot plot, with red dots for the things that are bright in the mCherry channel;

(left) the mCherry brightness histograms.

 

Rainbow beads histogram at the bottom (brightest beads are just a bit off-scale on the high end).

 

If any of you have an opinion, please let me know which flow rate you think is ‘best.’ I’m thinking either 25 uL/min or 100 uL/min.

 

Any other thoughts/critiques are also welcome. We’re planning to run real data next week.

 

Thanks-

Preliminary Attune Data.Dec 8 2016.pdf

Ariel Haim Hecht

unread,
Dec 9, 2016, 2:24:18 PM12/9/16
to sbsc-flow...@googlegroups.com
Interesting. Do you know what your mean events/sec at the different flow rates?

During our Attune training the rep suggested that bacteria be run on the slower rates, either 12.5 or 25 ul/min. FWIW, I’ve always seemed to be getting the best results on 12.5 ul/min, ~10,000 events/sec. I think it’s very interesting how the shape of the gated population on the FSC-SSC plot changes at the different flow rates. There seems to be a sort of bulge appearing at the bottom of the gated population that increases in size with the flow rates. Any thoughts on what that might represent? Could it be cell doublets?

I’m also intersted in how the shape and position of the mCherry histogram changes as a function of flow rate. To my eye it looks like the peak shifts rightward a bit and broadens with increasing flow rate. Do you know if that’s in fact happening? Might be interesting to see mean, median and mode of the gated population as a function of flow rate.

-Ariel

-- 
You received this message because you are subscribed to the Google Groups "SBSC Flow Cytometry Working Group" group.
To unsubscribe from this group and stop receiving emails from it, send an email to sbsc-flow-cytom...@googlegroups.com.
To post to this group, send email to sbsc-flow...@googlegroups.com.
Visit this group at https://groups.google.com/group/sbsc-flow-cytometry.
To view this discussion on the web visit https://groups.google.com/d/msgid/sbsc-flow-cytometry/BY1PR09MB04536F323765770434344765E7870%40BY1PR09MB0453.namprd09.prod.outlook.com.
For more options, visit https://groups.google.com/d/optout.
<Preliminary Attune Data.Dec 8 2016.pdf>

Nicholas DeLateur

unread,
Dec 9, 2016, 3:16:52 PM12/9/16
to sbsc-flow...@googlegroups.com
Everyone at the SBC runs their bacteria at our lowest setting, at 30 uL / minute. 
We also have a white paper on how low density leads to better data. This is attached. 
I don't recommend going above 5000 events/second, but I'm completely ignorant how much that is machine dependent. 

To check for double cell events I've seen people look at FSC vs SSC but I'm ignorant to the exact protocol, the idea is that if you get two populations, one of them is doubles. I'll look more into the specifics, the need for this is normally ablated by proper dilution and slow rates. 

Can you speak more to the "crystals" you see? 

Nicholas 

I have reproduced your other email here so it can be kept more in one place, if that's OK. 

"I got my first day in with the Attune flow cytometer yesterday. Did some practice runs with the E. coli strains Nick sent.

 

Some initial thoughts on measurement assurance steps that I would recommend adding to any protocols:

 

1.       Run buffer, focusing fluid, and media blanks. M9 media, in particular has the potential to form crystals that are similar in size/scattering to E. coli.

I would actually recommend running a media blank through the entire growth, dilution, and sample preparation process. So for the ‘Nielsen’ protocol we’re running: add one or more M9 media blanks to the plate for the day-0 overnight culture; then carry those media blanks through the entire process as if they were real samples, and measure them on the flow cytometer to check for issues with sample cross-contamination and/or other particulate. Also, before the 20x dilution into PBS, run a PBS blank on the flow cytometer, as a quality control check.

2.       Measure the sample carryover. As part of the protocol development for each instrument, after each E. coli sample, run a focusing fluid blank sample to check for carryover. Determine what rinsing protocol between samples is adequate.

Ariel- with the Attune instrument, running tube samples, it looks like running ‘Sanitize SIP’ 2 times between samples is a decent starting point for eliminating carryover"


On Fri, Dec 9, 2016 at 2:24 PM, Ariel Haim Hecht <arih...@stanford.edu> wrote:
Interesting. Do you know what your mean events/sec at the different flow rates?

During our Attune training the rep suggested that bacteria be run on the slower rates, either 12.5 or 25 ul/min. FWIW, I’ve always seemed to be getting the best results on 12.5 ul/min, ~10,000 events/sec. I think it’s very interesting how the shape of the gated population on the FSC-SSC plot changes at the different flow rates. There seems to be a sort of bulge appearing at the bottom of the gated population that increases in size with the flow rates. Any thoughts on what that might represent? Could it be cell doublets?

I’m also intersted in how the shape and position of the mCherry histogram changes as a function of flow rate. To my eye it looks like the peak shifts rightward a bit and broadens with increasing flow rate. Do you know if that’s in fact happening? Might be interesting to see mean, median and mode of the gated population as a function of flow rate.

-Ariel
On Dec 9, 2016, at 11:13 AM, Ross, David J. (Fed) <david...@nist.gov> wrote:

Here’s some example data I got off the Attune yesterday at different flow rates.
Data is for the mCherry only E. coli strain that Nick sent.
Growth protocol was shorted for initial test: grow overnight, dilute 300x, grow 3 hours, put on ice, dilute 20x for measurement.
 
For each flow rate, I’ve attached plots of: 
(left) the SSC vs. FSC as a density plot; 
(middle) the SSC vs. FSC as a dot plot, with red dots for the things that are bright in the mCherry channel;
(left) the mCherry brightness histograms.
 
Rainbow beads histogram at the bottom (brightest beads are just a bit off-scale on the high end).
 
If any of you have an opinion, please let me know which flow rate you think is ‘best.’ I’m thinking either 25 uL/min or 100 uL/min.
 
Any other thoughts/critiques are also welcome. We’re planning to run real data next week.
 
Thanks-

-- 
You received this message because you are subscribed to the Google Groups "SBSC Flow Cytometry Working Group" group.
To unsubscribe from this group and stop receiving emails from it, send an email to sbsc-flow-cytometry+unsub...@googlegroups.com.
To post to this group, send email to sbsc-flow-cytometry@googlegroups.com.
<Preliminary Attune Data.Dec 8 2016.pdf>

--
You received this message because you are subscribed to the Google Groups "SBSC Flow Cytometry Working Group" group.
To unsubscribe from this group and stop receiving emails from it, send an email to sbsc-flow-cytometry+unsub...@googlegroups.com.
To post to this group, send email to sbsc-flow-cytometry@googlegroups.com.

Visit this group at https://groups.google.com/group/sbsc-flow-cytometry.
To view this discussion on the web visit https://groups.google.com/d/msgid/sbsc-flow-cytometry/8E4CEAD8-5291-4857-B2D4-13AEF83FB809%40stanford.edu.
Effects of Sample Dilution.docx

John Sexton

unread,
Dec 9, 2016, 4:08:04 PM12/9/16
to sbsc-flow...@googlegroups.com
Hmmm, very interesting, thanks for sharing David. Is there any chance you could upload the raw FCS files to the Google Drive? Maybe a new folder in `2016_mini_interlab\data\ecoli` which includes the date measured (YYYYMMDD) in the folder name?

Rainbow beads histogram at the bottom (brightest beads are just a bit off-scale on the high end).

At what flow rate were the beads measured? Might be interesting to see how beads behave across flow rates.

Might be interesting to see mean, median and mode of the gated population as a function of flow rate.

I agree with Ariel that it would be interesting to see how appropriate summarizing statistics change as a function of flow rate.

-John


Ross, David J. (Fed)

unread,
Dec 9, 2016, 7:32:39 PM12/9/16
to sbsc-flow...@googlegroups.com

Additional info on our preliminary data, as requested:


Summary stats (mean, median, mode) are from the gate around the bright mCherry population (see previous post). I'm a little disturbed that they don't even have an option for geometric mean as a summary stat.

Number of events and event count rates are also for mCherry bright events. Since we had the thresholds set pretty low, there were a decent number of additional events (~200/s or so).


My very preliminary guess is that the extra bulge in the scatter plots at higher flow rates (below the main cell population on the FSC vs. SSC plots, FSC~10^4) is from small spherical cells (~1 um in diameter) whereas the main population is from the more normal rod-shaped cells. Not sure how to check this or whether we care.


Beads plot I showed was at 100 uL/min. Other flow rates look very similar. I didn't have time to check for a similar trend as in the mCherry-E. coli


Crystals are apparent on the FSC vs. SSC plot for some M9 media blanks I ran (with M9 that had been chilled and not filtered). I'll send an example plot on Monday. Those events partially overlap the main region of E. coli cells - but with a different shape on the plot.




From: sexton...@gmail.com <sexton...@gmail.com> on behalf of John Sexton <john.t...@rice.edu>
Sent: Friday, December 9, 2016 4:07 PM
To: sbsc-flow...@googlegroups.com
Subject: Re: [sbsc-fc] Preliminary data
 
To unsubscribe from this group and stop receiving emails from it, send an email to sbsc-flow-cytom...@googlegroups.com.
To post to this group, send email to sbsc-flow...@googlegroups.com.
mCherry stats.pdf

Ariel Haim Hecht

unread,
Dec 10, 2016, 1:04:45 AM12/10/16
to sbsc-flow...@googlegroups.com
Very interesting, thanks for sending Dave!

Did you order your M9 pre-made or did you prepare it yourself?


For more options, visit https://groups.google.com/d/optout.
<mCherry stats.pdf>

Ross, David J. (Fed)

unread,
Dec 10, 2016, 8:58:44 AM12/10/16
to sbsc-flow...@googlegroups.com

We're mixing our own M9. Pre-mixed was out of stock when we went shopping. From stories I've been hearing (something about a PhD thesis on auxotrophy gone bad, years of work wasted), that's a good thing.


In addition, I'm looking for a sensible and not too expensive QC solution for our media. I think Adam Arkin said his lab is now running some kind of mass spec analysis on all culture media. Even if we trust Sigma-Aldrich and make our own M9, the cas-aminos are a potential source of random problems.



From: sbsc-flow...@googlegroups.com <sbsc-flow...@googlegroups.com> on behalf of Ariel Haim Hecht <arih...@stanford.edu>
Sent: Saturday, December 10, 2016 1:04 AM
Reply all
Reply to author
Forward
0 new messages