-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1
Hello Gregoire-- > > I'm using a SA protocol on DNA. I run this protocol with xplor 1.2.1 and > xplor 2.14, and I noticed that the final result is somewhat different. > After looking at the output, I'm thinking that the difference comes from > velocity assignement (minimization does exactly the same in both version). > Here's my questions : > > 1) What is the last version that xplor 2.16 is compatible with (without > any protocol modification)? I assume you refer to the version of Xplor-NIH. This became decoupled from XPLOR at XPLOR version 3.851. All XPLOR 3.851 scripts should *run* with the latest version of Xplor-NIH. Please let us know of any exceptions. Likewise, older Xplor-NIH python scripts should work properly, unless otherwise noted. Now, the answers may change. We reserve the right to fix bugs, and thus change the results. Also, due to finite numerical machine precision, precise results will almost always differ when comparing results computed on different architectures. > > 2) Is there any 'rules' to upgrade script from old xplor version to > newer version? > The best/only way to exactly reproduce (within floating precision) old results, is to use the old version of Xplor-NIH. Then one should ask the question: whay do they differ. If you use your own parameters, results will not change unless we've fixed one or more bugs. The distributed parameters do change, however. Our policy for distributed parameters is that changes to topology or parameter files will involve a version number change. The latest and most tested topology and parameters will be available via the names nucleic.top, nucleic.par, protein.top and protein.par. Best regards-- Charles -----BEGIN PGP SIGNATURE----- Version: GnuPG v1.4.6 (GNU/Linux) Comment: Processed by Mailcrypt 3.5.8 <http://mailcrypt.sourceforge.net/> iD8DBQFF5H1FPK2zrJwS/lYRAugSAJ4k2zUZRO4mlCqcF9yFBdjyigKgJACeP5Pp K9oxLROUlV+Wn3TKaqdfFzg= =hzIk -----END PGP SIGNATURE-----
