1. The Compute utility allows users to compute and visualize global PTS distributions for their own traits of interest.


2. Users will need a collection of trait-associated SNPs, corresponding their trait of interest, in order to calculate and visualize PTS global distributions.


3. Users should organize their trait-specific SNP sets in a tab delimited text file (the SNP table) with the following fields:
1. SNP – the SNP rsID from dbSNP.
2. Effect Allele – the specific allele {A, T, C, or G} that influences the expression of the trait of interest.
3. Phenotype – a name or descriptor for the trait (phenotype) of interest. Note that users can include multiple traits for simultaneous PTS calculation from a single SNP table
4. Effect size – the odds ratio (OR) or β-value that measures the magnitude of the effect for that SNP. If effect sizes are not available, users can leave this position empty, which will default to an unweighted PTS.


4. An example SNP table is provided. The example table can be loaded to the server for a test computation, and/or it can be downloaded so that users can visually inspect the required format.


5. Once the user SNP table is prepared:
1. For users with < 1000 trait-associated SNPs:
   Paste or upload your SNP table in the form shown on the page To paste click on the upper left cell of the SNP table form and then paste.
2. For users with >= 1000 trait-associated SNPs:
   Upload your SNP table and provide an email to get your results

Users should pay close attention to quality control (QC) issues when assembling collections of trait-associated SNPs for PTS calculation. The PTS scores the GADGET computes are very much dependent upon the quality and accuracy of the input data. Specific QC issues that users may want to consider include, but may not be limited to:

• correct (positive) SNP strand orientation
• SNP strand ambiguity
• low genotyping accuracy for indels
• correct dbSNP build (150)
• linkage disequilibrium (LD)