Missing study information

[mightyphil2000]

There are a growing number of datasets in Open GWAS that have no information on their study of origin. This makes the datasets difficult to use.

For example, below is a study of fresh fruit intake where the pmid, doi and consortium fields are missing.

ao<-ieugwasr::gwasinfo()
 data.frame(ao[which(ao$author=="Lyon")[1] , ])
        id              trait note group_name year consortium author
1 ieu-b-45 Fresh fruit intake   NA     public 2020         NA   Lyon
                sex population unit nsnp sample_size       build subcategory
1 Males and Females   European   NA   NA      222722 HG19/GRCh37 Behavioural
     category ontology ncase ncontrol  doi pmid mr priority sd
1 Risk factor       NA    NA       NA <NA>   NA  1        0 NA

When a study is unpublished and has no pmid or doi, I think it should be a minimum requirement to enter a study name in the consortium field and/or to provide a link to documentation. Can we add a field called url, where we can link to documentation in google docs or word online? We can eventually convert the url to a doi, using the Bristol research repository. e.g. along lines of https://data.bris.ac.uk/data/dataset/aed0u12w0ede20olb0m77p4b9

It seems most of the datasets with no information on their study of origin are from UK Biobank, FinnGen and Biobank Japan. I have started using the y$api_metadata_edit() function from the GwasDataImport package to edit the consortium fields for these studies so that it is clear where the datasets come from. We should think about adding a study field to the meta data or think about changing the consortium field to study, which is more generic.