# LinkML and Valid Inputs - mmaSet column

**URL:** <https://odm.discourse.group/t/linkml-and-valid-inputs-mmaset-column/207>\
**Category:** New features\
**Tags:** odm-v3\
**Created:** [August 14, 2026, 7:02pm UTC](https://odm.discourse.group/t/linkml-and-valid-inputs-mmaset-column/207 "2026-08-14T19:02:50Z")\
**Posts on this page:** 1\
**Page:** 1

<div class="post-metadata">

**Author:** ![matthomson](https://avatars.discourse-cdn.com/v4/letter/m/dec6dc/32.png) [@matthomson](https://odm.discourse.group/u/matthomson)\
**Post date:** [August 14, 2026, 7:02pm UTC](https://odm.discourse.group/t/linkml-and-valid-inputs-mmaset-column/207/1 "2026-08-14T19:02:50Z")

</div>

Within the dictionary, categorical attributes and measures/methods have their range of inputs/ennums defined in the `mmaSet` column, which specifies with set (specifically a part with `mmaSets` part type, with defined membership in the `sets` table) defines the valid inputs for that field.

There are, however, some fields that rather than taking a value defined in a particular `set`, take any part with a specific `partType` value - think `measure` in the `measures` table, which can take any part where `partType == measurements`.

When generating the linkML however, these input relationships are not clear. To work around this, I am thinking of amending the `mmaSet` column to take _either_ an mmaSets part, _or_ a partType, to make this more explicit.

Any thoughts? @dmanuel @jeandavidt ?