In the original metadata table, each row is a study with potentially multiple EFO mappings represented as a single, comma-separated string value. We extract mappings from the original metadata onto a separate table where each row is an individual mapping. However, currently the labels are copied over as they are found in the original metadata. Instead, we should look up the labels for each CURIE and use that label instead of the original one.
In the original metadata table, each row is a study with potentially multiple EFO mappings represented as a single, comma-separated string value. We extract mappings from the original metadata onto a separate table where each row is an individual mapping. However, currently the labels are copied over as they are found in the original metadata. Instead, we should look up the labels for each CURIE and use that label instead of the original one.