Use of beads as viability reference controls

[claude-chew]

Hi @DrCytometer, Have been using version 1.5.6 and really liking the improvements. Quick question, I have been using beads for some of my viability controls, but the control file has to specify them as 'cells' in order to move forward. I've just tricked it by defining them as cells and letting the gating and spectra extraction do its thing and it has seemed to work just fine.

Does this matter as long as it can gate and extract spectra for unmixing? Will there be support for using beads a viability reference controls in the future?

Thanks,
Claude

[DrCytometer]

It shouldn't matter as long as you're happy with the results. I'm hoping to put in place a major overhaul of the whole first part of the workflow to remove the gating entirely and use a more automated approach, which is being developed and driven by Nathan Laniewski here:
https://github.com/nlaniewski/spectracle/tree/master
Might be a bit before I integrate this.

This hasn't been announced yet, but if you want to beta test, AutoSpectral's control cleaning and per-cell AF extraction methods have been put into Honeychrome:
https://honeychrome.cytkit.com/
This is still a work in progress, and will also be adjusted with the upcoming overhaul, but it is already moving down that path. The per-cell fluorophore optimization part of the unmixing has not been incorporated yet.

[claude-chew]

Ok thanks for that clarification, I just didn't know if there was some magical thing happening in the background depending on whether control type was bead or a cell.

Yes, I have honeychrome installed, but just haven't had the time to play with it, most likely a post-Cyto project

Thanks for continuing to maintain and upgrade AutoSpectral!

[DrCytometer]

There is some magic in the background, yes, but it pertains to the use of the matching unstained sample. If you use the internal negative, everything will be fine whether your sample is mapped to "beads" or "cells". If you use a separate unstained sample, you'll only be allowed to apply cell-type unstained samples to cell-type single-stains, and vice-versa, which would cause a bunch of small problems down the line. So, if you need to trick the setup, either use the internal negative for mis-typed samples or include a matching mis-typed unstained sample (e.g., unstained beads that you call cells).

[claude-chew]

Thanks for the acknowledgement by the way, I'm honored!

[claude-chew]

@DrCytometer thank you for including the spectral_qc_report with the normalized spectra against the library reference, I was trying to do that manually using flowSpectrum