Using GnuPG to Decrypt Restricted-Access Data

Since 1 January 2011, all restricted-access data available from PhysioNet have been provided to authorized users in unencrypted form via PhysioNetWorks, so the procedure described on this page is needed only to decrypt files that were downloaded in 2009 and 2010.

Restricted-access data available from PhysioNet were formerly provided as AES-encrypted files. These notes demonstrate how to decrypt these files, using a short text file as an example.

1. If you have not previously done so, install GnuPG (gpg). This is a standard component of most GNU/Linux distributions; versions are also freely available for Mac OS X and MS-Windows.
2. Download sample.key.
3. Download sample.txt.gpg.
4. In a terminal window, run the command:

         gpg --passphrase-file sample.key sample.txt.gpg
   

   This command decrypts sample.txt.gpg using sample.key, and writes the decrypted contents into a new file (sample.txt). Some newer versions of gpg may require an additional --batch option, like this:

         gpg --batch --passphrase-file sample.key sample.txt.gpg
   

5. To confirm that you have decrypted sample.txt.gpg correctly, compare sample.txt with original.txt.

To verify that the procedure above works as described, we tested a number of versions of GnuPG. On Mac OS X 10.5, we tested GNU Privacy Guard version 1.4.9. On MS-Windows XP, we tested gpg4win-1.1.4 (which installs gpg 1.4.9), as well as Cygwin's gpnupg 1.4.9-2. On Fedora and Ubuntu GNU/Linux, we tested gpg 1.4.9 and several later and earlier versions. Any version produced after October 2000 should work.

Please be aware that encrypted files are also compressed. For example, the decrypted and uncompressed sample.txt is 674K, nearly eleven times the size of sample.txt.gpg (62K).

 Name                    Last modified      Size  Description
 Parent Directory                             -   
 original.txt            06-Apr-2009 03:52  674K  
 sample.key              06-Apr-2009 05:23   43   
 sample.txt.gpg          06-Apr-2009 03:52   62K