../
|
reconstruct/
|
992
(9 years ago)
by jmduarteg:
Removed unneeded code
|
modelit/
|
1301
(8 years ago)
by hstehr:
overwriting default contactTypes.dat with the one needed for CMView release versions (only contains standard contact types). The old version with all esoteric contact types is renamed to contactTypes.all.dat
|
scripts/
|
1633
(6 years ago)
by jmduarteg:
Removing uniprot update script, doesn't belong in owl repo
|
jars/
|
1647
(6 years ago)
by jmduarteg:
Fixing bug in jaligner: the NeedlemanWunschGotoh method used by owl.core.sequence.PairwiseSequenceAlignment had a bug causing infinite hanging when using very long sequences. The issue would happen when either of the 2 sequences to align was longer than 32768 (16 bit) and the other sequence would match it after the 32768 region, a test case is included in this commit to demonstrate it. The problem was an overflow: an array used for the traceback procedure was declared as short[] (16 bit), changing it to int[] (32 bit) the issue disappears.
As the development in jaligner seems to be stopped I've simply downloaded the source, fixed the issue there (plus removing the loggers) and repackaged in a jar (jaligner-bugfixed.jar) that includes the sources.
|
src/
|
1688
(6 years ago)
by jmduarteg:
New default for DEF_NUM_CELLS: 12. After a scan of the whole PDB (Oct 2013) the worst case we found was 4jgc having a contact on the 11th cell.
|
gpl.txt
|
974
(9 years ago)
by duarte:
Adding license files
|
LICENSE
|
974
(9 years ago)
by duarte:
Adding license files
|
.project
|
1638
(6 years ago)
by jmduarteg:
Changing default eclipse project name to owl
|
build.xml
|
1644
(6 years ago)
by kumaran:
1)Javaversion variable is added in build.xml to change the version easily.
2)Pymol exited with error status 139 due to very long selection string written in the form of every residue(3+4+5+8+9+10). It is fixed by writing in the form of (3-5+8-10)
|
.classpath
|
1647
(6 years ago)
by jmduarteg:
Fixing bug in jaligner: the NeedlemanWunschGotoh method used by owl.core.sequence.PairwiseSequenceAlignment had a bug causing infinite hanging when using very long sequences. The issue would happen when either of the 2 sequences to align was longer than 32768 (16 bit) and the other sequence would match it after the 32768 region, a test case is included in this commit to demonstrate it. The problem was an overflow: an array used for the traceback procedure was declared as short[] (16 bit), changing it to int[] (32 bit) the issue disappears.
As the development in jaligner seems to be stopped I've simply downloaded the source, fixed the issue there (plus removing the loggers) and repackaged in a jar (jaligner-bugfixed.jar) that includes the sources.
|