15th January 2019: Please note, the SumHer Server is not yet finished. Keep checking back!
For the most accurate results, you should run SumHer manually, using LDAK to calculate a tagging file containing only SNPs for which you have summary statistics. However, we also provide a web version which uses a pre-computed tagging file. This tagging file is constructed using a generic list of SNPs, and so the accuracy of the web version depends on for how many of these generic SNPs you have summary statistics.
At present, the web version can only be used to estimate the SNP heritability of a trait, the amount of confounding bias, and heritability enrichments for the 24 functional categories used by Finucane et al. Therefore, to estimate heritability enrichments for other annotations, or genetic correlations between pairs of traits, you must instead run SumHer manually.
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _
Prior to using the web version, you must format your summary statistics. Your file must have the headers Predictor, A1, A2, Direction, Stat (or P) and n. We recommend predictor names are in the format Chr:BP (e.g., 12:345678), but you can also use rs numbers. Below is an example script, or for more detailed instructions see Summary Statistics. Your file can be saved as a text file (preferred extension .txt), zipped (with extension .zip) or gzipped (with extension .gz). Note that to save space, you can first download our generic SNP list, then reduce your summary statistics to these.
Once you have formatted summary statistics, click here to use the SumHer Server.
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _
