Finding proxy SNPs in the outcome data when using the "read_outcome_data" function.

[nicholaspudjihartono]

Please make sure that this is a feature request! If you have questions about how to use TwoSampleMR please use the Discussions function instead.

Let's say I have "SNP1" in my exposure data, but "SNP1" is not present in my local outcome data (i.e., the outcome GWAS was not taken from the "available_outcomes()" function). As the outcome data is local, we cannot use the "extract_outcome_data" function to automatically find proxy SNPs. Instead, you have to use the "read_outcome_data" function, which currently does not have the option to automatically search for proxy SNPs.

Therefore, I wonder if there is a way to find proxy SNPs in the outcome data using the "read_outcome_data" function , or maybe this feature should be implemented inside the "read_outcome_data" function.

This is crucial because some diseases like Juvenile Idiopathic Arthritis, do not have available GWAS with substantial number of samples in the IEU GWAS database (which is where the "available_outcomes()" function get their data). So I prefer to download the summary statistics of a GWAS that I prefer from the GWAS catalog. However doing this means that I cannot use the "extract_outcome_data" function which handles proxy SNP automatically.

[wangyiyisheng]

Did you solve this problem？

[laleoarrow]

This function was not implemented yet in the TSMR package now. I contributed my code addressing this issue in a recent PR (#519). I hope it may be of some help to you. Though, the current version could potentially have some stability concerns. Any advice or feedback would be welcomed.