modelBasedQCPlots

[sara n]

Hello all,

I was wondering if there is a way to test the normality other than using the normal Q-Q plots, such as  shapiro.test(data) , or exploring the r^2 for example for a single protein, not just exploring the significance of a Protein from the p-value and adj.pvalue?

---------------------------

Also, say I want to explore a list of Proteins instead of which.Protein="all"

here is what i have and it worked when i manually type it in but 

dataProcessPlots(data=quantData,

                 type="QCPlot",

                 which.Protein = c("sp|A0A075B6J9|LV218_HUMAN"

                                   ,'sp|A0A075B6K2|LV312_HUMAN'

                                   ,"sp|A0A075B6K4|LV310_HUMAN"),

                 address="")

i want it to work from a list:      my_list1 <- list(data.frame$Protein)

dataProcessPlots(data=quantData,

                 type="QCPlot",

                 which.Protein = my_list1,

                 address="")

Error in datafeature$PROTEIN %in% temp.name : 

  object 'temp.name' not found

[Mateusz Staniak]

Hi,

what is the purpose of using a list?

Kind regards
Mateusz

[Mateusz Staniak]

Hi,

then using a vector (as in your first example) is OK. To get significant proteins, do for example:

x <- <groupComparison output>$ComparisonResult

sig_prots <- unique(x[x$adj.pvalue < 0.05, "Protein"])

and pass this new vector to which.Protein

Kind regards

Mateusz

[Mateusz Staniak]

Hi,

this doesn't look like an error from the current version of MSstats. Can you please try with the latest version? Let's see if it works.

Kind regards

Mateusz

[sara n]

Hello, 

After installing the new update and adjusting to the new changes,  Unfortunately I get this new error. May I see your example R.script to see what might be the problem with mine?

> x <- testResultOneComparison$ComparisonResult
> sig_prots <- unique(x[x$adj.pvalue < 0.5, "Protein"])

> # normal quantile-quantile plots
> modelBasedQCPlots(data=testResultOneComparison, type="QQPlots",
+                   which.Protein = sig_prots
+                                       )
Error in getSelectedProteins(which.Protein, all_proteins) :
  object 'selected_proteins' not found

[Mateusz Staniak]

Hi,

I just ran this code:

library(MSstats)
example("modelBasedQCPlots", package = "MSstats")
unique(testResultOneComparison$ComparisonResult$Protein)
sig_prots = c("IDHC", "PMG2")
modelBasedQCPlots(data=testResultOneComparison, type="QQPlots", address="",
                  which.Protein = sig_prots)

and I didn't get that error. It just produced the qq-plots. Does that help? Please let me know if the error persists.

Kind regards

Mateusz