You have uploaded at some point protocol data to a run but haven’t imported data to all readout definitions.
Alternatively, you might have created a new readout definition in your protocol after you already have imported data into runs.
In either case, you want to complement existing runs with the missing/new data.
Just like compound batch level data, you cannot overwrite existing data in your readout definitions via the file (bulk) import. To change ("overwrite") existing values, you would have to manually edit the data fields within your vault; for bulk import though, you can complement missing data, but with a slight caveat:
Possibility 1: You want to add the missing readout definition and make these addition obvious when looking at the data in your table. The result would look something this, with the new readout in a separate row (Run view):
To obtain this, after uploading and “Add Readouts (protocol data) only”, map your IDs as usual, select the existing Run (Date) and map only the missing data column! Should you map all columns, you will end up with unwanted replicates!
Note, that calculations (e.g. an average) won't be affected by this entry having "two rows".
Possibility 2: The more likely scenario: You want the new data integrated seamlessly, displaying only one row in the search view:
This is currently no possible via "updating" the run per se: You will have to delete the runs entirely and re-import them again. See here on how to delete a run. If you registered with plates, you can, but don't have to, delete those imported plates. In other words, this is optional and you save yourself some steps by not deleting the plates from the initial (to be corrected) run. See here on how to delete a plate.