Page 2 of 2

Re: Acceptable / typical background levels

PostPosted: Wed Jul 31, 2019 8:52 am
by jimbomahoney
EDIT - The below was conjecture / misunderstanding on my part, but I'll leave the post as-is.

Another thought that I had after discussing with Fluidigm was the LCT (lower convolution threshold).

The default is 400 (DCs I assume).

Therefore, this corresponds to solution mode as follows:

(1000/13) * 400 = ~31k DCs.

This is because one push is 13 ┬Ás, whereas the solution mode acquisition is an "accumulation" over 1 second.

Therefore, if contaminants are present higher than 31k DCs in solution mode, this may impact the detection of weak events by pushing the LCT up and hiding them.

Thoughts / comments appreciated as always!

Re: Acceptable / typical background levels

PostPosted: Wed Jul 31, 2019 3:30 pm
by mleipold
Hi James,

I'm not sure that's completely correct.

The LCT is a summary across *all* channels. This is one reason why you may want to increase the LCT for high-parameter experiments or decrease it for low-parameter experiments. Even in a well-stained sample, each channel will have a certain amount of background, which can add up.

Unless you're having streaking in all or most channels or *terrible* background in just one, I haven't seen much issue with cell event determination (ie, Event Rate going really high as it thinks every bit of debris is an event, or even worse, Event Rate drops really low as it starts thinking you have *REALLY LONG* cells since the between-event data never drops below the LCT). I *think* this is because you usually have at least a few really bright markers (such as Ir) that helps drive the true Events' signal well above the LCT.

If anything, I've had at *least* as many problems with event rates and resolution when users run their samples too fast (even if the sample is "clean").


Re: Acceptable / typical background levels

PostPosted: Thu Aug 01, 2019 8:22 am
by jimbomahoney
Thanks Mike.

Yes, I chatted again with Fluidigm, who explained that the LCT units are not pulses or DCs - it's more of an internal / software unit, so it's not as simple as I had understood it to be.

And yes, I understand now that the LCT is getting pushed up by Ir anyway. I'll edit my original post too.