If your datasets/files are located elsewhere, e.g., in DAMS or another SI server, you can still get a DOI for them by pointing to them from a Metadata only record
After you click the + Create a new item button, check the box next to  Metadata record only and begin adding your descriptive metadata.

Describe your data

In order to publish, you will need to fill in all the required metadata fields.

1. Title - Please compose a meaningful and intelligible title.
   If the data are for a specific paper/article published elsewhere, preface the title with Dataset: followed by either part or all of the title of the paper.
    Fighsare will automatically use the filename of the first file you upload as the title. ick!
2. Authors - figshare automatically populates the author field with your information when you (the item owner) create an item, however, you can delete your name if you are uploading for another person. Dataset author is unrelated to ≠ dataset owner.
   you can add authors by typing in their names or pasting in an ORCID. If your co-authors are not using ORCIDs in their Figshare profiles, please give them a hard time about it.
   1. typing in the author name does a search of the entire Figshare author corpus. Be careful to confirm that the name you choose is in fact your co-author!
   2. If you're uploading for someone else, simply add the correct authors, then delete your name.
3. Category - choose at least one subject term. Figshare uses the Australian and New Zealand Standard Research Classification (ANZSRC) Fields of Research (FOR) codes. Start typing in the term you want and the field will auto-populate.
4. Keywords - up to you.
5. Description - you can reuse your data abstract, if you have one, here. If you don't have an abstract, now is a great time to write one!
6. License - the default for SI is CC BY NC 4.0, which is appropriate for SI assets (images, video) or datasets that have been analyzed and contain significant intellectual effort besides just the raw data.
   1. Feel free to use the CC0 license if your data is raw or the data is from a (US) Federally funded project or if you (or the primary author) is a Federal employee.
   2. If you are uploading code, please choose the Apache 2.0 license.
      
      

More about the "Category" field

More about Licenses