Internal steps of the browser when you load a page

From genomewiki
Jump to: navigation, search
  • When you load a page, your webbroser will send the current URL and your local cookie to Apache
  • The Apache webserver will execute the hgTracks binary, sending it the parameters from the URL
  • The hgTracks binary will first try find your user configuration ("cart") from the hgcentral.userDb (?) table, using the hgsid communicated either via the cookie or a hgsid parameter
  • hgTracks will first parse the hg.conf file in the cgi-bin directory which tells it how to open the configuration files, where to get track information from (all the specified trackDb tables) and how to login the mysql databse
  • It will then display the page based on your user configuration and the parameters specified, e.g. the position can come from the parameters, the activated/deactivated tracks will usually come from your cart
  • For each activated track it will get the track configuration data from one of the trackDb tables.
    • Local mirrors will usually have one standard trackDb from UCSC and one local one like trackDb_local or trackDb_John
  • Depending on trackDb, the track data will usually come from a table in the mysql database, sometimes also from textfiles specified in the track information (biwig, bigbed, etc)
  • The bin-field in a table (or the index of bigwig/bigbed) is used to restrict the number of features to the current window of the genome