FAQ  •  Register  •  Login

Extracting each push from IMD

Forum rules
Please be as geeky as possible. Reference, reference, reference.
Also, please note that this is a mixed bag of math-gurus and mathematically challenged, so choose your words wisely :-)
<<

rgrenfell

Contributor

Posts: 30

Joined: Thu Aug 27, 2015 12:31 pm

Post Thu May 20, 2021 3:49 pm

Extracting each push from IMD

Hi everyone, hope you are all staying safe

Bit of a weird question this, but has anyone looked at trying to extract individual push data from the IMD?

What I would like is to get a few thousand pushes, with the ion counts per channel, as a text file (or csv)

We have some samples where a metal is not retained in the cell very well, so shows a significant amount in solution between cells. Processing an IMD (using a CD marker to identify events), only gives us half the picture, and excludes the signal between cells. I'm worried that what we are seeing is just the metal in solution that happened to be passing at the same time as a cell, rather than the metal that is associated with the cell

I would like to see if individual push data could show us that there is, or isn't, a corresponding increase in metal signal during pushes containing a cell.

Reprocessing of the IMD within the Fluidigm software doesn't seem to be able to be granular enough to turn individual pushes into individual events. If it could export an FCS of single pushes as events, then plotting CDmarker ion-count vs time and overlaying metal ion-count might be a way I can correllate the two

I'm thinking that taking the rainplot and turning it into a csv file is what I need, but I don't know if there is a way of getting that out of the IMD
(and yes, I do realise its not going to be a small file!)

Hope this question makes sense!
Thanks
Richard
<<

BjornZ

Contributor

Posts: 43

Joined: Fri Jul 10, 2015 1:04 am

Post Thu May 20, 2021 4:14 pm

Re: Extracting each push from IMD

Rachel Finck has MatLab code for reading IMD files here: https://github.com/nolanlab/MatlabCytof ... master/imd. The file format is essentially a block of uint16 integers for the pushes and an XML descriptor.

Best,
Zach
<<

mleipold

Guru

Posts: 5866

Joined: Fri Nov 01, 2013 5:30 pm

Location: Stanford HIMC, CA, USA

Post Thu May 20, 2021 4:31 pm

Re: Extracting each push from IMD

Hi Richard,

Yes, you can convert an IMD to TXT.

Assuming you have a Helios: to get an IMD in TXT form, you have to into Service Mode (normally you run in User or Administrator mode). Then, in Process, there will be a new "Convert IMD to Text" button. The new TXT will be roughly the same size as your original IMD file.


Regarding metal streaking: yes, if a Cell Event is triggered by other channels, the Event gets written with that Time Interval across all channels. In other words, the signal intensity in the Streaky Channel will be taken as Cell-Associated even if it's Streak (non-cell-associated). So, you could have "background" intensity on your cells coming from a streaky channel.

The overall effect of the Streaking on your data somewhat depends on the intensity of the Streaking. In your Template, Noise Reduction is turned on by default. Fluidigm won't tell us how this works exactly, but it's roughly a background subtraction ( viewtopic.php?f=4&t=582&p=1885&hilit=noise+reduction#p1885 ). In general, if there's a small amount of streaking, this seems to get removed efficiently, but at some unspecified point, the Streak-Associated background starts to increase on your Negative population.


One other thing you can try: years ago, Rachel Finck created something called "Rainmaker" ( https://github.com/nolanlab/MatlabCytofUtilities ; https://github.com/nolanlab/MatlabCytof ... 2bf29abcdb ) which basically allows you to mimic the Rain Plot on snapshots of an existing IMD file.

Note: it's not being actively maintained (last update was 2015), and there's not really a Wiki to instruct. But it does exist.


Mike

EDIT: Zach beat me to Rainmaker. :)
<<

rgrenfell

Contributor

Posts: 30

Joined: Thu Aug 27, 2015 12:31 pm

Post Fri May 21, 2021 3:27 pm

Re: Extracting each push from IMD

Thanks both

I think the noise reduction was causing us a problem, by mostly removing what we are looking for.
We'll also take a look at the matlab script along with trying to access service mode.
Yes, its a Helios we're using

Thanks again & have a good weekend everyone

Richard
<<

mleipold

Guru

Posts: 5866

Joined: Fri Nov 01, 2013 5:30 pm

Location: Stanford HIMC, CA, USA

Post Fri May 21, 2021 3:30 pm

Re: Extracting each push from IMD

Hi Richard,

You can always reprocess the IMD to remove Noise Reduction, and see the effect on your data.

And the EQ beads can often be used as a nice "background detector" for non-Bead channels. In this case, you could see what the Streaking Channel looks like on the EQ beads +/- noise reduction.


Mike

Return to CyTOF data analysis

Who is online

Users browsing this forum: AdsBot [Google] and 3 guests