`adjustment` not applied when metadata used as `var` in `dittoSeq` and friends

[j-andrews7]

This is in the dev version, so Bioc 3.20:

example(importDittoBulk, echo = FALSE)
myRNA

# Basic dittoplot, with jitter behind a vlnplot (looks better with more cells)
dittoPlot(object = myRNA, var = "score2", group.by = "timepoint", adjustment = NULL)

Looks identical with z-score adjustment (or relative.to.max).

dittoPlot(object = myRNA, var = "score2", group.by = "timepoint", adjustment = "z-score")

Works fine for genes though:

dittoPlot(object = myRNA, var = "gene1", group.by = "timepoint", adjustment = NULL)

dittoPlot(object = myRNA, var = "gene1", group.by = "timepoint", adjustment = "z-score")

Note that it is still applied to genes when both metadata variable and genes are provided, e.g.:

multi_dittoPlot(myRNA,
                vars = c("score1", "score2", "gene3", "gene4"),
                group.by = "clustering", adjustment = "z-score")

Probably won't have any time to chase this one myself for at least a few weeks.

[dtm2451]

Not a deep one... it's just how it has always been. The docs for adjustment in ?dittoPlot:

When plotting gene expression / feature counts, should that data be used directly (default) or should it be adjusted to be
  - "z-score": scaled with the scale() function to produce a relative-to-mean z-score representation
  - "relative.to.max": divided by the maximum expression value to give percent of max values between [0,1]

[j-andrews7]

Huh, I totally thought it used to apply to metadata too. Well, that's on me then.

…

On Tue, Sep 24, 2024, 9:09 AM Daniel Bunis ***@***.***> wrote: Not a deep one... it's just how it has always been. The docs for adjustment in ?dittoPlot: When plotting gene expression / feature counts, should that data be used directly (default) or should it be adjusted to be - "z-score": scaled with the scale() function to produce a relative-to-mean z-score representation - "relative.to.max": divided by the maximum expression value to give percent of max values between [0,1] — Reply to this email directly, view it on GitHub <#156 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/ACOAQNGYR6QTERUUTNJHRV3ZYFXDJAVCNFSM6AAAAABOXBB6I6VHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZDGNZRGQYTMMJSGA> . You are receiving this because you authored the thread.Message ID: ***@***.***>

[dtm2451]

Easy enough to update.

Most of the machinery was put in place a while ago, but I stopped short of the final connection from an uncertainty of whether to make the change + my tendency to not change default behaviors without adding a level of control to recreate old behavior.

Just gotta think a bit on if an extra control should be added and if that control should default to 'on' or 'off'.

[cstrlln]

Just checking, I'm not seeing a difference when using z-score as adjustment in dittodimhex. relative.to.max is working thought, so not sure best way to compose.