# Metabolite prediction using 16S taxonomic abundance data

**URL:** <https://forum.biobakery.org/t/metabolite-prediction-using-16s-taxonomic-abundance-data/140>\
**Category:** MelonnPan\
**Created:** [January 14, 2020, 7:02pm UTC](https://forum.biobakery.org/t/metabolite-prediction-using-16s-taxonomic-abundance-data/140 "2020-01-14T19:02:10Z")\
**Posts on this page:** 2\
**Page:** 1

<div class="post-metadata">

**Author:** ![dpak283](https://yyz1.discourse-cdn.com/flex027/user_avatar/forum.biobakery.org/dpak283/32/161_2.png) [@dpak283](https://forum.biobakery.org/u/dpak283)\
**Post date:** [January 14, 2020, 7:02pm UTC](https://forum.biobakery.org/t/metabolite-prediction-using-16s-taxonomic-abundance-data/140/1 "2020-01-14T19:02:10Z")

</div>

Hello there,

Is MelonnPan accepting taxonomic level abundance tables as input files yet? I am getting an ID mismatch error between training and test set when I try this. Does the input have to be UniRef90 annotated gene families?

Thanks.  
-Deepak

---

<div class="post-metadata">

**Author:** ![franzosa](https://yyz1.discourse-cdn.com/flex027/user_avatar/forum.biobakery.org/franzosa/32/3511_2.png) [@franzosa](https://forum.biobakery.org/u/franzosa)\
**Post date:** [January 16, 2020, 5:50pm UTC](https://forum.biobakery.org/t/metabolite-prediction-using-16s-taxonomic-abundance-data/140/2 "2020-01-16T17:50:30Z")

</div>

The features do not have to be UniRef90s – we do some experiments training on 16S OTUs in the paper. It sounds to me like this might be an error matching the gene/taxon profiles to the metabolite profiles (which need to be 1:1)?