Allow user-defined `sumstatsColHeaders` mapping as argument

[albert-ying]

Hi, just a suggestion: I think pre-defined column-name mapping is convenient but not very flexible. For example, I have a summary stats table with a1 as effective allele and a0 as other allele. The format_sumstats will, however, treat a1 as reference allele and a0 as effective allele. I'll have to change the colnames of the file manually or update the sumstatsColHeaders just for this dataset to get it to work.

I think it would be better if the format_sumstats could take colnames mapping as the argument. For the example I showed, it would be much better if I can just do something like:

format_sumstats(path = raw_file, A1 = "a0", A2 = "a1", ...)

or

new_col_header = data.frame(Uncorrected = c("a0", "a1"), Corrected = c("A1", "A2"))
format_sumstats(path = raw_file, colheaders = new_col_header)

Thank you!

[Al-Murphy]

Hi @albert-ying,

Thank you for your suggestion! While we tried to make the mapping file all encompassing for summary statistics file headers and how people interpret them, I agree that it would be a good option to let the user input their own if required. To this end, I have included a mapping_file parameter where the user can input their own mapping file in the same format as data(sumstatsColHeaders). This will be available in 1.1.6+ which should go live today, we are just incorporating a few more changes to this release.

Thanks, Alan.

[Al-Murphy]

v1.1.7 has been pushed to Github with this functionality in place. The functionality will propagate to Bioconductor in the next release around September.