Handle aggregate epidemiology data (and add SIR)

[anngvu]

There isn't a lot of coverage for analysis outputs because that is usually not prioritized for annotation (since analysis content will usually appear in the publication). However, we have a study where the main product is analyses producing statistical tables of NF incidence, etc. This is derived data from the original individual-level government/hospital registry data that the investigator has access but can't be uploaded to Synapse. So they would like to upload the data, where the first use case asks for adding standardized incidence ratio (see https://sagebionetworks.jira.com/issues/NFOSI-1032), which would have parent class dataType population-based statistical data at least. Such data can still be used for further epidemiology analysis.