There's a recurring pattern in many of the PEPs we're creating: 1. `append` an attribute (arbitrary string) for every sample 2. use this string to `derive` a desired value See [this PEP](https://github.com/refgenie/refgenomes.databio.org/blob/c2d5e1348e4277291e6cb6b87d8ded58945d2f68/asset_pep/refgenie_build_cfg.yaml#L9-L18), for example Therefore it seems like adding a `sample_modifiers.append_derive` would be useful