Congrats, Geoffroy! Looking forward to reading it 🙂

Sending emails != having email conversations. Big difference in workload. The lvl of communication requested was unclear in the paper. Anyway, as per your other reply, such a mess is fully avoided by only using data already mentioned in a paper (I hope they'll update publication tags reliably ...)

It would help if the demand for communication is automatically dropped from the DRI as soon as a journal publication is linked to the data set in question, otherwise the future of applied bioinformatics is an administrative nightmare.

If under the pressure of publish-or-perish the DRI that demands communication becomes universally used, the amount of communication to just do a simple genome mining task (e.g., how many strains of x can do y because of gene z) explodes - still worrying because it's time and energy consuming.