In the mzTab-M files exported from MS-DIAL 5.5, the retention_time_in_seconds column in the SMF section appears to contain minutes, not seconds.

I tested it with an example file from HUPO-PSI/mzTab-M/examples:

• msdial_5.5.251021GUI_Height_zenodo14263441.mztab (link)

I checked it in R (using rmzTabM):

install_github("lifs-tools/rmzTabM")
library(rmzTabM)
tbl <- readMzTab("/msdial_5.5.251021GUI_Height_zenodo14263441.mztab")
smf <- extractSmallMoleculeFeatures(tbl)
rt <- as.numeric(smf$retention_time_in_seconds)
max(rt)

Output:

>   max(rt)
[1] 7.515305

This seems implausible for a typical LC-MS experiment if interpreted as seconds, but is reasonable if interpreted as minutes. I also repeated this check on the other MS-DIAL example files in HUPO-PSI/mzTab-M/examples and on an additional mzTab-M file generated with MS-DIAL, and observed similar retention-time ranges.

Based on this, it looks like the values in retention_time_in_seconds might currently be exported in minutes rather than seconds. Could you please check whether the mzTab-M exporter is using minutes here, and if so adjust the export so that retention_time_in_seconds is reported in seconds?