88e5bf615b39ac9a8c93c140dbac9c1f5ea12885
max
  Wed Nov 26 03:07:29 2014 -0800
extending makefile refs #14393
diff --git src/beacon/README.txt src/beacon/README.txt
new file mode 100644
index 0000000..fedf636
--- /dev/null
+++ src/beacon/README.txt
@@ -0,0 +1,31 @@
+The GA4H beacon system http://ga4gh.org/#/beacon is a tiny little webservice
+that accepts a chromosome position and replies with "yes", "no" or "maybe"
+if it finds information in a data track. This is to allow queries from outside
+users on protected variants. If they find variants in someone's beacon, they
+then can contact the institution to get more information about it.
+
+It's implemented as a CGI script (query) which uses bigBedToBed and a set of
+.bb files.  No mysql is required and the data files live in the same directory
+as the script and the bigBedtoBed binary. The makefile copies everything
+together into cgi-bin/beacon/.
+
+The data for the beacon, the .bb files, are copied into the cgi-bin directory
+from /hive/data/genomes/hg19/bed/beacon
+
+We don't have personal data at UCSC so the beacon in this directory serves
+primarily LOVD and HGMD, which we cannot distribute.  These tracks are also
+not in the table browser, but the beacon makes it possible to at least query
+the positions in some way. 
+
+To get usage info, run the CGI with no option, e.g.
+http://hgwdev-max.soe.ucsc.edu/cgi-bin/beacon/query
+
+The CGI keeps a list of counts of IP+date, to limit the maximum number of
+queries to 1000 a day. This is to make it harder to query the whole genome and
+identify a person from the responses. This limit can most likely be increased.
+
+The CGI also accepts INSDC accessions instead of "chr1" or just "1". The
+mapping is stored in insdcToUcsc.tab. 
+
+The CGI currently only supports hg19.
+