FAQ  •  Register  •  Login

Time axis cut off: >60% events below zero

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 :-)
<<

marthabrainard1

Contributor

Posts: 20

Joined: Wed Aug 25, 2021 2:10 pm

Post Wed Aug 25, 2021 5:31 pm

Time axis cut off: >60% events below zero

Hi

Has anyone ever seen a cut off time axis? I am losing >60% of events. I de-barcoded this file with one other file and the time v event length plot on this file looks normal. The time v event length plot looks normal on the concatenated file too. I don't have the option to upload an attachment below this thread (I just see options and poll creation buttons), but the histogram has a peak at 0 if you go the customize axis button
<<

mleipold

Guru

Posts: 5796

Joined: Fri Nov 01, 2013 5:30 pm

Location: Stanford HIMC, CA, USA

Post Wed Aug 25, 2021 7:56 pm

Re: Time axis cut off: >60% events below zero

HI Martha,

I'm not sure what you mean about poll creation. There should be an Options tab, and then an Upload Attachment tab. If there's not, please email me off Cytoforum and we can try to take a look at your account.

You can also post the image to Google Drive or some other online site and post the link.


What program are you using to look at your file? FlowJo? Cytobank? FCSExpress? etc?

One possibility: sometimes in the file processing (Normalization, concatenation, cutting off a bad part of the file and exporting a new FCS file, etc), the Time dimension can get corrupted. In most cases, this wipes out the Time values (ie, Time parameter/column still exists, but it's Empty/Blank)

In FlowJo, the way this usually appears is that the Time axis will look ok (not piled up on 0), but if you look at the axis tick-marks, there will be "0" and then no further values (like 1000K, 2000K, etc).


Honestly, I haven't found a good way to fix this. There is some information about how long a Time increment is in CyTOF (eg, viewtopic.php?f=3&t=1073&p=4324&hilit=TIMESTEP#p4324), but formally I don't know of any package that *fixes* an emptied Time parameter.


In a practical sense: it doesn't stop you gating on other parameters. And while I know that a messed up Time parameter affects the MATLAB/Finck normalizer, I don't think it affects Premessa normalizer.


Mike
<<

marthabrainard1

Contributor

Posts: 20

Joined: Wed Aug 25, 2021 2:10 pm

Post Wed Aug 25, 2021 9:48 pm

Re: Time axis cut off: >60% events below zero

Capture_time_issue_x.PNG
Capture_time_issue_x.PNG (42.95 KiB) Viewed 4663 times


Capture_time_issue.PNG
Capture_time_issue.PNG (65.93 KiB) Viewed 4663 times


Hi Mike,

Thank you. The attachments upload button was here when I clicked reply! I am using Flowjo, here are screenshots.
<<

mleipold

Guru

Posts: 5796

Joined: Fri Nov 01, 2013 5:30 pm

Location: Stanford HIMC, CA, USA

Post Wed Aug 25, 2021 10:01 pm

Re: Time axis cut off: >60% events below zero

Hi Martha,

Thanks for the plots.

I'm not completely sure I understand your experiment when you say " de-barcoded this file with one other file and the time v event length plot on this file looks normal. The time v event length plot looks normal on the concatenated file too."

Could you walk us through your entire experiment (staining, barcoding, running, norm, concat, debarcoding, etc)? Just so we see all the steps of the sample and file processing in case there's something that jumps out.


In short: I understand your previous post to mean that you had 2 Raw files which you then (normd? and) Concatenated, then debarcoded into 2 Debarcoded Sample files where File 1 has a wonky Time dimension and File 2 has a normal Time dimension.

Is that correct?


Mike

Return to CyTOF data analysis

Who is online

Users browsing this forum: No registered users and 9 guests