Missing values even when MBimpute = TRUE

[LRY]

Hi, 

I am using MSstats v4.6.0 on a DDA dataset, but even when I let MBimpute = TRUE, the summarized protein level data still have a lot of missing values:

```proteinGroups <- read.table("proteinGroups.txt", sep = "\t", header = TRUE)

infile <- read.table("evidence.txt", sep = "\t", header = TRUE)


# D = 10 mol UPS1, C = 25 mol
annotation <- read.table("summary.txt", sep = "\t", header = TRUE) |>
  select(1:2) |>
  slice(1:6) |>
  rename(Run = Raw.file) |>
  mutate(Condition = str_sub(Experiment, end = 1)) |>
  mutate(BioReplicate = as.numeric(as.factor(Condition)))


quant <- MaxQtoMSstatsFormat(evidence = infile,
                             annotation = annotation,
                             proteinGroups = proteinGroups)

dat <- dataProcess(quant,
                   normalization = FALSE,
                   summaryMethod = "TMP",
                   censoredInt = "NA",
                   MBimpute = TRUE,
                   maxQuantileforCensored = 0.999)

prot <- select(dat$ProteinLevelData, RUN, LogIntensities, Protein) |>
  pivot_wider(names_from = RUN, values_from = LogIntensities) |>
  column_to_rownames("Protein")
anyNA(prot)
```

The link to the dataset is here, and the MaxQuant output is in `Ratio2.5_txt.zip` file. 

Any suggestion is very appreciated. 

Cheers

L

[Devon Kohler]

Hi L,

Sorry for the delayed response, I just came across this post and wanted to write a response in case anyone else has this question. MSstats will only impute values in certain situations. It assumes values are missing due to reasons of low abundance and requires at least one other feature that we can use to determine what value should be imputed. I've attached an image below which shows how imputation is performed.

TLDR: MSstats will not automatically impute all values and many will be left blank. We only impute values if there is enough information to impute them well.