Will Shaw

Dupuytren's disease & drug repurposing

Data and downloads

Every file behind this site, free to download and reuse with credit.

Why it’s usefulIf you want to check my work or build on it, it's all here, with notes on the few columns that are easy to misread.

Everything in one file

Download all data (.zip)

All CSVs, the six research reports, the seven structure files and the GWAS figure. Data as of 31 October 2026.

Spotted an error?

Corrections make this resource better. Please say which record and what the source says.

Suggest a correction

Or email willshawphd@gmail.com

Files

FileContentsRows

How to cite

Shaw, W. (2026). Dupuytren's disease and drug repurposing: an open evidence map [Data set and website]. Data as of 31 October 2026. https://willshaw.io/dupuytrens/
BibTeX 
@misc{shaw2026dupuytrens,
  author       = {Shaw, Will},
  title        = {Dupuytren's disease and drug repurposing: an open evidence map},
  year         = {2026},
  note         = {Data as of 31 October 2026. Built with Claude Science},
  howpublished = {\url{https://willshaw.io/dupuytrens/}}
}

Please also cite the primary studies you rely on; every row carries its PMID or DOI. Licence: text and compiled data CC BY 4.0. Source papers remain under their publishers' terms.

Before you reuse the data

A few things in these files are easy to misread:

  1. To count drug studies, use the studies_a_drug column. The treatment_class column files drug-plus-surgery trials under surgery, so counting by class misses them.
  2. "Not stated" in a row with abstract_available = FALSE means the information is missing. It doesn't mean the study found nothing.
  3. In the genetics chart, height shows how confident a finding is. It doesn't show how much a variant raises risk.
  4. Two trial-planning numbers are blank on purpose: the smallest change in nodule hardness that matters to patients, and how often untreated disease progresses. Nobody knows them yet, so any trial size built on them is a guess.
  5. Only one structure (3WD5) shows a drug bound to its target. The AlphaFold files are predictions of a protein's shape.

Fields were extracted from abstracts by a language model under a forced schema, then audited (see Methods). Treat the tables as an index into the literature, not a substitute for it.