Metaphlan2 with RNA data

Hi everyone,
We have a dataset of saliva metagenome and metatranscriptome and we ran metaphlan2 on both DNA and RNA reads. The output results seem fine and we want to compare the relative abundance of the taxa identified on both strategies.
We’re aware that Metaphlan may not be appropriate for this, because of the problem of marker genes used for the reference database and the copy number issue it may cause in the analysis. Has anyone ever tried this approach? Using Metaphlan on RNA data? Do you recommend any kind of normalization to deal with the data and make it comparable to the DNA? If not, do you recommend a different strategy?