Assign replications to multiple references

How do we deal with published reports from individual studies - e.g. this ML5 case:
https://journals.sagepub.com/doi/full/10.1177/2515245920917334

They should be linked to both so that we can show all originals for ML5 - but also link the specific replication publication to the original.

This also relates to preprints, which we should be able to link to their targets without duplicates in database.

We might need a "also published as" column for that - but need to remember that this is not symmetric. To be discussed and fixed soon, as it has implications across all interfaces (including pubpeer comments)