If you select 'Remote database' make sure credentials are defined in environment variables

Path to the tabular file to load

Type of the feature property (cvterm will be created if it doesn't exist)

Type of the target features in sequence ontology (will speed up loading if specified)

Match features using their name instead of their uniquename

Use this if you want to delay the job execution until some data is already loaded. The selected dataset will not be used for anything else.