Detritus Wiki

Procedure 1

• First we took the every provided file (listTEST5.txt, listTEST6.txt, listTEST15.txt, listTEST16.txt) and look for the proxies for every marker at http://www.broadinstitute.org/mpg/snap/ldsearch.php with,

    SNP data set: 1000 Genomes Pilot 1
    r2 threshold: 0.7
    Population panel: CEU
    Distance limit: 100
    

• We substitute every marker for all its proxies and construct every possible pair for each file
• Then we choose the common pairs to all files and construct a single one (p4.txt)
• This file must be converted in plink format (just put a column after the other)
• We run plink over each database looking for epistasis,

    plink --bfile bed_DB --epistasis --epi1 1 --epi2 1 --set-test --set sets.txt --allow-no-sex --out bed_DB_sets
    

• Now we extract from the results just the pairs corresponding to p4.txt for each DB.
• Finally, following Combining probability from independent tests: the weighted Z-method is superior to Fisher’s approach, M. C. WHITLOCK, doi: 10.1111/j.1420-9101.2005.00917.x, we calculate the combined p-value,

$$ p = \phi \left( \frac{1}{\sqrt{k}} \sum_{i=1}^{k} \phi^{-1}(p_i) \right) $$ where $ \phi $ is the Normal distribution cumulative function.