MetaboAnalystR - subsetting/filtering from metadata


I have LCMS concentration data where samples are organised by rows. I have corresponding metadata to these samples.

With MetaboAnalyst in R I can create the object with both of these inputs. With the server-based interface, I can successfully input the csv files. However, there are some functions with the data object that I’m wondering if anyone can provide some feedback on or insight…

With R, is there a code I can follow to create the object with the LCMS data, metadata, AND subset from this original object? The subset would be based on condition/s in the metadata e.g. Male/Female, Control/Disease, Disease with or without a certain comorbidity. This would be similar to Phyloseq where prune_taxa, prune_samples, subset_taxa, and subset_samples can be used to filter/subset the Phyloseq object. The Phyloseq object is very similar in concept to the Metaboanalyst object in that they are there for “data management”. Does MetaboAnalyst R or server have something similar?

With the webserver, the one-factor and metadata table avenues, is there any way to get the combination of these two platforms together? Also if the metadata input (currently, it can only be viewed) can be edited and used to filter the object of interest? It would save having to import and export many different Excel csv prepared outside of the Metaboanalyst package platform. For example, the sample group in one-factor analysis for the LCMS data can be assigned male or female in the Group column, but another another way of looking at the LCMS data for the same samples would be to assign Control/Disease in the group column.

Thank you.

Excellent suggestion! This is clearly doable via either MetaboAnalyst web or the R package. We are busy with MS2 features at the moment. This will be the next priority in our to-do list. @jess.ewald @Qiang

1 Like