BioJava dependency updated to latest bugfix version 4.2.9
Fixed problems where CLI was crashing in assembly creation in some edge cases.
NGL now displays all subassemblies of disjoint assemblies.
A few bug fixes
3.0.2 - 18th September 2017
BioJava dependency updated to latest bugfix version 4.2.8
A few bug fixes
3.0.1 - 21st April 2017
Assembly probability scores now shown in assemblies table. Default sorting by probability.
Fixed issue with large heteromeric assemblies not being described correctly.
Known issues: DNA/RNA-only assemblies are not called correctly, disjoint assemblies are not shown correctly in thumbnails and 3D view.
3.0.0 - 16th March 2017
Full assembly enumeration and prediction of most likely assembly
Probabilistic scoring of interfaces and assemblies, including confidence assignments.
2D and 3D visualizations of the lattice graph
WebGL JavaScript viewer NGL
Surface sequence entropy 3D visualizations in browser, no need of PSE files
Based on BioJava 4.2, replacing OWL
NCS operators now correctly used to build viral capsid structures
Large structures with multiletter chain ids now supported
PDB biounit assignments now displayed side-by-side with assembly predictions
Interface clustering now based on contact similarity instead of RMSD
Known issues: DNA/RNA-only assemblies are not called correctly, large heteromeric assemblies don't work correctly.
2.1.2 - 19th November 2014
XML files now have residue details
No similar structure search button will appear if sequence clusters are empty for the chain
Source code unified into a single repository with submodules
GXT updated to 3.1.1
Some bug fixes
2.1.1 - 21st August 2014
Downloadable command-line program is now correctly packaged
New database statistics page
Similar structures search is now a lot faster
Some bug fixes
2.1.0 - 8th July 2014
Jmol will launch now in JS mode if Java is not detected
Predictions now also done at interface cluster level by averaging within cluster
XML data downloads
New feature: searching similar structures from sequence clusters
Disulfide bonds are no longer used as a bio call criterium
Complete redesign of data model behind the scenes
Many bug fixes
2.0.5 - 31st January 2014
Now Phenix mmCIF files will be properly parsed
Jmol version updated to 13.0.18
Better warnings for low resolution, high R-free and EM structures
Problems with alignment window are now solved in IE
A few bug fixes
2.0.4 - 17th January 2014
Major user interface change: upgraded to GXT 3.0.1, new top navigation bar, improved "My Jobs" panel, nicer fonts and much more
New homologs information pop-up window providing details on the identity distribution and taxonomy of each homolog used in the MSA
Introduced interface clustering. A button on top of the interfaces table now triggers a cluster view of equivalent interfaces
New FAQ and Publications pages. Improved and expanded Help page
2.0.3 - 27th November 2013
Improvements in top panel: now split into experimental info panel and sequence info panel. Also some improvements in residue details window: scores are now better displayed
Cleaned up the input page: only alignment parameters can now be modified in advanced input
Issues with windows and tooltips displaying badly in IE are now gone thanks to GXT 2.3
Warnings for low resolution or high R-free added
Interfaces below 35 Å2 are not displayed anymore
Fixed issue in interfaces calculation: the interface search extends now to more distant cells, missing interfaces in structures such as 1was will now appear
A few bug fixes
2.0.2 - 22nd August 2013
Precomputed EPPIC results are now available for the whole PDB, after typing a PDB code the results will appear immediately
Now operators displayed as symmetry icons with the algebraic operators in tooltips. Operators colored in red to mark infinite interfaces
A few bug fixes. In particular issue with crashes in parsing PDB files having no TER records (or badly placed ones) is now fixed
The static pages are now fully crawlable by search engines
Some style improvements
2.0.1 - 19th April 2013
Fixed displaying problems in residue details table, the ugly white frames disappeared
Now the homologs panel in top panel contains links to download the colored-by-entropies PSE files produced by the EPPIC command line program. Very useful to get a visual impression of the conservation of residues at the surface of the structure. This shows graphically why EPPIC works
Better handling of UniProt reference mapping, especially for multi-segment mappings
More warnings added: unreliability of surface residues (surface residues not aligning to UniProt reference), if too many then NOPRED call is given
Display and style improvements
A few bug fixes
2.0.0 - 15th February 2013
First full public release
Cofactors of more than 40 non-Hydrogen atoms are now taken into account. They will be considered as part of the protomers in ASA calculations. This affects especially predictions in small proteins with large cofactors like hemoglobin
PDB residue numbers are now used for warning messages and output jmol, pdb and pse files
The missing eppic.conf file is now provided in downloadable eppic.zip
The guessing of file types is now more flexible, it will allow PDB files starting with: HEADER, TITLE, COMPND, SOURCE, EXPDTA, REMARK, DBREF, SEQRES, CRYST1, MODEL, HELIX, SHEET. Any other file not starting with one of these lines will be rejected
Blast results are now precomputed for all PDB entries which makes runs for deposited PDBs a lot faster
The web session will now expire after 24 hours of inactivity. The "My Jobs" list will then be lost
Style improvements
Many bug fixes, especially in relation to parsing PDB and mmCIF files