paper-qa icon indicating copy to clipboard operation
paper-qa copied to clipboard

Supporting Supplemental Information

Open jamesbraza opened this issue 4 months ago • 3 comments

Background

Sometimes a paper's supplemental information (SI) shows up in a separate document. The SI shares the same DOI as the main text. Ideally the title in the SI mentions the term "Supplemental", but as shown in DOI 10.1021/acscatal.1c04879 at https://pubs.acs.org/doi/10.1021/acscatal.1c04879:

  • The main text and SI have the same title
  • The SI's title page does not contain the word "supplemental"

I can't find an authoritative MLA citation guide on how to specifically cite an SI, but the APA 7th Referencing Guide mentions the citation should include the text "Suppl.".

Current State

Here is the inferred citation of paper-qa==5.25.0's MLA citation prompt: https://github.com/Future-House/paper-qa/blob/v5.25.0/paperqa/prompts.py#L77-L83:

Text Inferred MLA Citation
Main Lu, Xiaoyun, et al. "Enzymatic DNA Synthesis by Engineering Terminal Deoxynucleotidyl Transferase." ACS Catalysis, vol. 12, no. 5, 2022, pp. 2988–2997. ACS Publications, https://pubs.acs.org/doi/10.1021/acscatal.1c05996. Accessed 23 July 2025.
SI Lu, Xiaoyun, et al. "Enzymatic DNA Synthesis by Engineering Terminal Deoxynucleotidyl Transferase." National Center of Technology Innovation for Synthetic Biology, Tianjin Institute of Industrial Biotechnology, Chinese Academy of Sciences, 2025. Accessed 2025.

Mainly, we can see the citation prompt did not recognize the SI as an SI. What happens next is, when we acquire complete paper metadata using providers such as Semantic Scholar or Crossref, the final DocDetails are actually identical (as shown in https://github.com/Future-House/paper-qa/issues/1005).

Feature Request

The request is for PaperQA to add first-class support for SI. This could entail implementations such as

  • Adding an Enum field to DocDetails for the "role" of the document: main text, supplemental information, protocol, etc.
  • Making a separate Doc subclass with fields tailored towards SI

jamesbraza avatar Jul 24 '25 18:07 jamesbraza

Thank you for mentioning this issue!

lucky0218 avatar Aug 09 '25 10:08 lucky0218

Shall I close my previous issue?

lucky0218 avatar Aug 09 '25 10:08 lucky0218

Hi @lucky0218 yes thanks again for the report. Lets keep the other issue open too, that issue will be closed in our next release. The solution I took for that issue is a breaking change, so we're going to do a major version bump for it.

jamesbraza avatar Aug 09 '25 18:08 jamesbraza