Updated make-aglappe script (and renamed to make-owl).
Changed aglappe strings to OWL (following new name of library)

Now analyseHistone will output the coordinates of the side chains of ARGs (NH1 atom) and LYSs (NZ atom) instead of the CA atoms.
Some new methods in Pdb

Moved Interval* and IntPair* classes to tools package

Added some useful tools to the default package from my workspace. More clean up.

Removed some obsolete classes.

Refactoring of inconsistently named methods in RIGraph

Added some new prediction measures for DecoyScoreSet: rank, delta rmsd, n% enrichment value
Extended the range of the distance bins.

Reorganised the project with a src folder for java source files.
Added a jars dir with all jars needed for the project.
Added .project and .classpath pointing to relative path of jars. 
The project should now work out of the box after a check-out with eclipse. No need to setup external jars or anything.

New rmsd method to calculate rmsd on residue intervals instead of whole chain.
Fixed bug: checkScop was throwing exception when it was called on a PdbfilePdb that didn't have a PDB code

Added new scoring method: atom distance based scoring (scoring based on multiple distance bins, instead of just on contacts)

Fixed bug (hopefully, didn't test!) introduced in last revision. Wasn't initialising properly the maps

Moved up to Scorer initResMap and initAtomMap

Delta Rank

merged tinker changes into trunk

DALI alignment correction

Fixed bug: was not handling correctly sequences with 'X' in reconstruction
Fixed bug: Pdb.getPhi/Psi were failing because hasCoordinates(int,String) was NullPointing for unobserved residues. 
Added test case for getAllPhiPsi

New convenience methods in MySQLConnection for simple retrieval of String, Integer and Double arrays from an SQL query. Some documentation clean-up.

updated constant LATEST_VERSION to CSA version 2.2.11 (of 7th August 2009)

tolerate unobserved residues in DALI alignments

Added DALI Alignment option

New classes for triplet scoring.

Now printing some more rmsd stats (numbers of structs below certain rmsd cutoffs).

Moved all decoy scoring stuff to its own package.
Split CombinedScorer into two classes AtomCombinedScorer and ResCombinedScorer. Made combined scorer more efficient: now loading only once the graphs.

Now writing rmsd ranges to stats files

New CombinedScorer class to combined type+count scoring. Now scoreDecoys doesn't take anymore a minSeqSep, it will be taken instead from the scoring matrices meta data.

Added method to get the scoring type

One more field in DecoyScoreSet: the decoy group to which the decoy set belongs. Writing it to db too.

Fixed bug: parse() was not closing file when throwing an exception. The bug surfaced when parsing a dir with a lot of decoys that were somehow malformed PDB files, sometimes it would throw a FileNotFoundException with message "Too many open files".

Now writing more complete info to score files. New feature: can also write decoy score results to database.

Increased number of neighbor-count bins. For higher cutoffs the number of bins go up dramatically.

New class DecoyScoreSetGroup to represent a set of DecoySets. Now scoreDecoys much clearer.

Got rid of redundant class DecoySetStats, its members were all just properties of DecoyScoreSet. Modified accordingly scoreDecoys.

New classes to score decoys based on contact counts (residue or atom contacts). New superclass for all scorer classes Scorer. 

New executable scoreDecoys to score sets of decoy sets. 

Added isAllAtom to Pdb and getNumHeavyAtoms to Pdb and Residue. Fixed some docs that were still not properly clarifying about hydrogens.
Added complete main methods to AtomTypeScorer and ResTypeScorer with command line options and defaults.

New classes for scoring protein structures based on residue/atom type empirical potentials.

Bug fixes in pdb file parser. Was throwing string out of bounds exception with some pdb files.

rewrote method getPerAtomDistances such that atom serials do not need to match between the two structures (matching corresponding atoms based in resser)

New executable writePerAtomDistancesToBFactor (see doc for details), new method Pdb.getPerAtomDistances(Pdb referencePdb), made Atom.DEFAULT_B_FACTOR public.

Major change: now hydrogens are read in Pdb and subclasses
New getChirality method in Residue.
Added some more unit tests.

Major commit: 
- change of the implementation of Pdb (interface remains backwards compatible). Now data stored in a proper object-oriented way. Pdb not abstract anymore, a protein model can be created by using one of the constructors
- added many tests and test files for Pdb (and merged some already existing tests from the default package)
- removed obsolete classes related to msd
- bug fix: PdbfilePdb now properly reads files with alt codes
- all tests passed

deleting obsolete package proteinstructure.features

moving all classes in proteinstructure.features to features

new method getUniprotTypeName()

updated static convenience methods for loading Pdb objects

adding main method for testing

adding compareTo method to compare two IntervalSets

New and updates feature types (UniprotFeature, PrositeFeature, CsaFeature)

New classes for scanning Prosite Database

HasFeatures.addFeature() now throws OverlappingFeatureException and new InvalidFeatureCooordinatesException

New package proteinstructure.features with basic interfaces and classes for a general feature mechanism which could eventuelly replace classes like Scop, ScopRegion, ECRegion, CatalyticSite, CatalyticSiteSet and the corresponding variables in the Pdb class.

New class Atom. Not doing anything yet. Now only need it so that Residue class can compile.

Refactored Residue->PolResidue. Now Residue will be the proper implementation of a residue as part of a Pdb

Minor changes: added comments, docs, reorganized

Extracted constants. Written a basic test for consurf parsing.

New check in PdbfilePdb parser. New test for pdb file parsing in PdbParsersTest.

Adapted PdbasePdb and CiffilePdb to new mmCIF format (introduced in new remediated release in Feb 2009). Now the atom_sites_alt table is gone so we have to read alt locs from atom_site instead. All tests passed (PdbParsersTest)

Fixed docs, now removing some temp files that weren't been removed

New methods copy() and translate()

Now aligning x-axis with reference residue ARG G42. Now theta values in [-pi,pi].

Now Polymer can read also DNA. analyseHistone now uses DNA as well: axis calculated with DNA, outputting z/theta stats for phosphates of DNA.

New classes to perform analysis of histone lateral surface.

updated documentation

New class PdbSet providing convenience methods to load sets of PDB objects from list files, directories, ...

Adding a symbol for a stop codon (*, STP)

New method to filter on minWeight

Fixed bug: copy() was throwing NullPointer when the RIGraph had a null SecondaryStructure annotation.

New method getMotifNoGaps

Changed the value of the unobserved residue constant from "?" to "X". This is to fix CMView that had a lot of problems with it.

Fixed bug (an old one!): the atomser2coord and atomser2resser maps contained also hydrogen atoms whilst the resset_atom2atomserial map contained only non-H atoms. Now all contain only non-H. The "bug" only surfaced after the rewrite of the Pdb.writeAtoms method: writing the atom lines would give a NullPointer when encountering a Hydrogen (if the read pdb entry contained Hyrdrogens).

updated to latest CSA version, added toString() methods

New static methods for conveniently loading PDB objects without needing to deal with exceptions (loadStructureOrExit, loadStructureOrNull)

New class GeometryTools for simple geometric operations (currently implemented: torsion angles)

Now Reconstructer can also take RIGraphs with "unobserved" residues (i.e. with missing nodes).
Model pdb files that Reconstructer creates now contain strictly only CA atoms.

New classes Reconstructer and ModelPdb. Reconstructer is a one-stop solution for reconstructing from contact maps (RIGraph) using BoundsSmoother and Embedder. At the moment doesn't support reconstruction of contact maps with non-observed residues.

added method getTypeDescription to output a string representation of the secondary structure type (currently not distinguishing between different numbers of states)

fixed bug in runAccess : temporary file could not be created when pdb code was not known, now using '1xxx' in this case

New enum class AminoAcid

added methods to set b-factor values; updated writeAtomLines to write b-factors if set (currently b-factors are not being read when creating a pdb object); added method getAtomSersFromResSer(); added writeAtomLines method with simpler interface

Fixed runNaccess to use unique file names and documented it. Wrote also a junit test case for it.

Fixed in BoundsSmoother: now the lower bound will be set to the triangle inequality value only if the newly found lower bound is bigger than the hard spheres bound.
Now we also introduce contiguous CA distance constraints (3.8)
In main now checking for violations to the full bounds matrix.
New method in Pdb for getting distance matrix as Jama object

New method in Pdb to set atoms for all coordinates of a given atom. Used in BoundsSmoother and Embedder mains.
Changed slightly how parameters are passed in BoundsSmoother.

white TO talyn

First implementation of the embedding part of the EMBED algorithm by Havel: from an exact complete distance matrix gives 3d coordinates for the embedding.

Got rid of the hardcoded mysql server, user, password. Now read from .my.cnf file

better error handling in AlignmentEvaluator; calling recompiled polypose15 in PolyposeRunner

removing debug output in moveToCenter() methods

added methods and executables to calculate 3d superimpositions using PolyposeRunner

added moveToOrigin() methods

added getGaplessColumns(start, end)

keep on evaluating common cores even if rmsd increased

Minor fix in mirror(): now using getters to access data

adding transform method to rotate/translate the structure

New method in BlastHit to transform the BlastHit local alignment into one that contains the full sequences of subject and query.

Fixed warnings and error handling in addSecStructAnnotation

Rewrote the secondary structure handling in Alignment. Also changed first parameter in writeFasta to be a PrintStream not an OutputStream

GraphAverager: got rid of redundant code in countVotes. Changed signature of one of the constructors: now requires target sequence.
averageGraph: adapted accordingly
Alignment: new method removeSequence (untested), fixed some docs

now outputing the gapless columns; now assuming that alignment file has axtension .fa (for extracting the name)

New static method in BlastHit to get subjectId from templateId
New member source in TemplateList

Fixed bug: PDB/graph/scop data loading issues were not correctly implemented. 

BlastHitList: new method getHit(subjectId), a new lookup HastMap member added for that.
TemplateList: new method get(i)
Sequence: new convenience method writeToFastaFile

Now cleaning up correctly paul temp files
Now closing tcoffee's log properly

Initial commit of StructAligner interface and implementation PaulStructAligner.
New reader in Alignment: now we can read CLUSTAL alignment files, not extensively tested. Also Alignment now only throws generic FileFormatErrors instead of specific file format errors.

New tests package (should be a folder and under it packages, but for now we can't do that until we move all source code to a src folder).
Moved all existing tests to appropiate tests packages.
New test PdbParsersTest: does cif against pdbase at the moment.
Adde equals method to SecStrucElement
Added some getters to Pdb

Created getters for missing fields in BlastHit. 
New method removeAll in TemplateList.

Committing empty class HashGrid for later implementation

New class AlignmentEvaluator to compare multiple structure alignments in terms of conserved core size and RMSD

Given a serialVersionUID to class to get rid of warning

committed by stehr: New methods getGaplessColumns and getLongestNonGappedRegion; fixed severe bug in old getGaplessColumns

can now superimpose subsets of residues + removed canExecute() call which was incompatble with Java 5

First version of PolyposeRunner (does an optiomal RMSD superimposition of multiple protein structures)

added comment

Switching to JUNG2-beta1:
- got rid of our custom fix to removeEdge
- now both addVertex and removeVertex are correctly implemented: they update properly the nodes2serials map. 

added missing javadoc comment

added methods to return the residue mappings for this alignment

New class RIGraphMinSpanningTree for getting the Minimum Spanning Tree of a RIGraph. This is simply a wrapper around JUNG's PrimMinimumSpanningTree.
New method setSerials2NodesMap in RIGraph.

Got rid to direct references to serials2nodes Map in write graph methods, now using the getSerials() and getNodeFromSerial() methods instead.

Fixed bug: was not initialising the type variable in BlastHit and GTGHit constructors

Now averageGraph can use PDB/CASP TS files as templates.
TemplateList/Template classes now can read PDB data either from pdbase database or from files.

Now considering also '.' (dots) part of the  FASTA header tags.

Changed the type used for the set of template graphs/pdbs from arrays to TemplateList. Now the loading of pdb data and getting the graphs is done by TemplateList.

-OligomericState class added

-Pdb
a.checkOligomeric and getOligomericState methods added
b.calcVolume, calcSurface and calcCompactnessCoefficient methods added

Fixed bugs in averageGraph:
- output of benchmark statistics and pairwise comparison was overlapping
- benchmark mode was broken after introduction of align mode (when no sequence file given we were always creating dummy sequence in both align or benchmark modes)

Improved:
- better handling of modes through mode variable that takes 1 of 3 states
- better input check
- in align mode now calling another GraphAverager constructor instead of creating dummy seq/tag within averageGraph
- changed constructor in GraphAverager: now sequence not a parameter, instead taken from alignment. Check for input moved down to averageGraph
- now allowing alignments that contain other sequences than the target and the templates
- new exception GraphAveragerError thrown in GraphAverager constructor when sequences of alignment and templates don't match

Pdb:
-pdbCode and chainCode info added in the error messages in runDssp

getNumScopRes method added in Scop class

Pdb
-restrictToScopDomains methods added to restrict to multiple scop domains
-restrictToIntervalSet is public now

now handling special cases that used to give NaN values

added option to load only first models; added new constructor taking a map of names to protein objects

New feature: now we can enforce omega trans conformation for the torsion angle of the peptide bond.
New method in Pdb: getOmega

added boolean parameter 'showSequence' to Alignment.writeSecStructMatching()

Added printing of alignment positions in writeSecStructMatching

chechScop now looks also for chain A in scop.cla when pdbChainCode is NULL

Now casp method string can be several lines.

Now enforcing that atom serials have to be in ascending order

Also adding caspParents member to graph classes, value will be kept from Pdb if present; adding method to GraphAverager to retrieve the most frequently used parents specified in the templates

Dummy commit (forgot comment for previous one): added field caspParents with getter and setter to Pdb. This field is filled when reading from and used when writing to from Casp TS files, therefore interface of writeToCaspTS file has changed slightly (parents now has to be set in Pdb object instead of passed to method as parameter)

New class PhiPsiAverager to get consensus of phi/psi angles from a TemplateList and an Alignment. Tested with a few examples and seems to work. The wrapping of angles at 180/-180 is not yet taking into account, i.e. if an interval falls in the region just below 180 and just above -180, no consensus will be found.
Pdb: added some checks to methods getPhi/Psi so that it doesn't fail when there's no coordinates.
Changed yet again the design of TemplateList/Template. Now loading of PDB data happens upon call of the loadPDBdata method. Changed dependencies accordingly.

Casp 3D predictions now written with NULL chain code

Now phi/psi angle methods using a more general getTorsionAngle. New method to get all phi/psi in a TreeMap

Added methods to calculate phi/psi angles

-writeAtomLines and writePdbFileHeaders have become public
-writeAtomLines and write[CaspTS|Pdbfile]Headers take now PrintStream argument instead of PrintWriter.

New feature: averageGraph now also writes out the CASP RR file (if -c specified) for the average graph at Cb 8.
Made a bit prettier the CASP RR output: now contacts sorted and weights with only 2 decimal figures.

added option to GraphAverager and averageGraph to create dummy sequence with length of alignment to view the overlaid graphs without mapping to a target sequence

Implemented parsing of PsiPred secondary structure prediction files in SecondaryStructure class. 
New member in SecondaryStructure: sequence. Constructor must pass the sequence now, changed all references.
New method in Alignment to print a "graphical" matching of secondary structures.

added 'title' member to Pdb and Template objects, currently only set when loading from Pdbase

Changed default gravity value and default IDENTITY_SCORE for similarity graph

Fixed bug in BlastUtils.writeClusterGraph: was not working at all. Wrong curly brackets, was still using a GDT cutoff instead of rmsd and was never getting Pdb correctly out of the Template.
Two new members in Template: GTGHit and BlastHit, so we can reference back to the hit if needed (like if we want to get the evalue/score). Got rid on MySQLConnection as a member, not very nice design, now a parameter for the specific methods that need to query for PDB data.

Got rid of the refined/unrefined field when writing CASP TS files. Now we should be complying correctly with the CASP 8 format.

Now writing the PDB ids for the PARENT fields of CASP TS files correctly (with underscore between pdb code and chain code)

Now also copying caspMethodStr and caspAuthorStr in AIGraph.
Updated java docs

Fixed bug: now try/catch of getPdbInfo moved down to getPdb and getScopSccsString (getScopSccsString was throwing NullPointer when pdb load failed)

Added code to be able to filter out template ids (pdb codes) by maximum release date.
Some improvements in Hit classes.

Added members and method to ProtStructGraph and Pdb to write AUTHOR and METHOD fields to Casp files. Added missing reconstructFast methods to TinkerRunner.

Added a few more methods to the Template classes: Template now behaves differently in that it doesn't load the PDB data on construction but on demand (when calling getPdb() or getScop())

Added methods to be able to print "graphical" output for GTGHits, all equivalent to the methods in BlastHit
Added compare() and writeIdsToFile() in TemplateList.

Initial commit of GTG parsing classes.
Added method in PdbasePdb to map observed residue sequence serials to internal (cif) residue serials

New method getObservedSequence

Fixed error output of PdbCodeNotFoundException

PdbfilePdb: catching previously unhandled NumberFormatException, added method setSequence() to override sequence read from file (needed for Casp prediction where 'real' sequence is known but residues may be missing in prediction)
RIGEnsemble: many changes

Extracted constant NULL_CHAIN_CODE that was still hard-coded in MANY places. Now should be safe to replace the value of the constant from "NULL" to something else.
Using TemplateList.readIdsListFile() in averageGraph, dumpseq and genGraph so that is now a standard way of reading list files with pdbcodes/chaincodes.
NOTE: this introduces an incompatibility with legacy list files: we can't read a NULL pdb chain code with the "NULL" string e.g. '1i1b NULL'. This is fine because since PDB-REMEDIATED all NULL chains are now "A".
Fixed a few things and made more standard the scripts dumpseq and genGraph

Introduced a new load method, so that is possible to specify cif chain codes (asym_ids) as input. Default is as before: input is pdbChainCode. NOT TESTED.

Moved readTemplatesFile method to TemplateList class (as a static)

Improved javadocs and comments

changed regexp for Casp header to allow target tag with lower case 't'

Added method writeToCaspTSFile. Outsourced writing the headers for PDB files and CASP TS files to other methods.
Refactored dumpseq to writeSeqToFile.
Improved a lot the java docs.

RIGraph:
-getCommonEdgesCount added

removed import which caused a warning

fixed one null pointer bug and some spelling mistakes

Added code to run psipred.
Fixed a few minor bugs.

New classes Template, TemplateList to be used in homology modelling pipeline.
Modified BlastUtils to use TemplateList class

write_graph_to_motiffile:
-node serials are assigned now only to non-orphan nodes

fixed bug in writing to gdl file (gravity 1 -> 1.0)

changed some parameters for writing GDL files

added method calculateSequenceIndependentMatrix()

RIGraph:
-bugs in write_graph_to_db methods concerning the node_id field have been fixed.

createGraphDb:
-the node_id in the single_model_node table changed to AUTO_INCREMENT

doClassifySingleModels:
-the interSS for CR added

RIGraph:
-write_graph_to_db methods treat the node_id field in the single_model_node table as AUTO_INCREMENT
-write_graph_to_motiffile method added

RIGraph:Now you can store unweighted graphs in db and also select the db that contains the ids for the single models.
-paramater weighted added to write_graph_to_db and write_graph_to_db_fast methods. If true, it works as before (atomWeight is written to db field weight) else the db field weight gets always value 1. Methods without this parameter still exist and work as before.
-protected field singleModelsDb and corresponding methods getSingleModelsDb and setSingleModelsDb have been added. All constructors set the singleModelsDb field to DEFAULT_SINGLEMODELS_DB which equals to "ioannis". Changes have been also made to write_graph_to_db and write_graph_to_db_fast methods to deal with the singleModelsDb field.

DbRIGraph, testDbRIGraph:Changes have been also made to address the reading of (un)weighted graphs from db.

doClassifySingleModels.sh has been added. You can create your own table with ids for the single models wherever you want.

createGraphDs.sh: mode to drop tables has been added.

reconstruct:
-rmsd to mirrored pdb has been added in the report file

Pdb:
-comment for outputing missing atoms added

PairwiseSequenceAlignment:
-constructor with parameters openScore and extendScore for the gaps has been added

Added comments and better javadoc for the get_graph methods

Correction in javadoc

Added new convenience method printSummary

RIGEnsemble doesn't extend ArrayList anymore, now the ensemble ArrayList is a member, so we now must use the RIGEnsemble interface exclusively

Fixed bug: was printing the value of accuracy for the coverage field.

Changed dumpseq: now complies with FASTA format (80 characters per line) and tag is now pdbCode+pdbChainCode (with no underscore)

New method writeToPaulFile

Added mirror() method in Pdb class
Added getComplement() method in RIGraph class

Now throwing exception in checkLengths().
Now fast checking of FASTA/PIR format (we check if first non-empty line starts with ">")

Fixed bug: now reading/writing #MODEL field in contact map files
In Pdb writeAtomLines now prints a nicer header with source info depending on source: file or db

changed header of contact map file from #AGLAPPE... to #CMVIEW...; when reading, both are accepted to ensure backwards compatibility

Fixed bug: write_graph_to_file now writes blank instead of "null" when pdbCode, pdbChainCode or chainCode are null.

Added constant NO_SEQ_SEP_VAL

New feature in GraphAverager: method to write the voters together with the edges; new constructor taking a RIGEnsemble
Fixed bug in RIGEnsemble: was not working for cif files.
Added a graph averaging output to main() in RIGEnsemble

Now forcing i<j in CASP RR files (as specified by the CASP RR format)

New constructor to read from simple graph files, i.e. list of contacts without any headers

Fixed bug: addEdgeIJ was not checking for presence of contacts

Added convenience methods: containsVertexI, addEdgeIJ

New methods getFisrt/Last residue serial

Drop sequence comparison in compare. Now we compare only length
New method getSize in RIGNbhood to be used instead of size that returns neighborhood size + 1

Fixed bug in compare(), was not checking correctly for sequence matching

Fixed reading od PDB files. We hope now to be catching all possible errors present in original PDB files, plus we read correctly CASP TS files:
- now reading (and requiring) TARGET record for CASP TS files
- now we always have (and require) a sequence (possibly with ? or X) for all cases: PdbfilePdb, CaspRRFileRIGraph, FileRIGraph.
- in PDB files we throw format exceptions for: insertion codes, residue numbers <=0, non-ascending order of residue numbers in atom lines, when residues of SEQRES seq and ATOM seq don't match. In addition we warn when starting residue in ATOM lines is >100
- we allow PDB files without a HEADER, i.e. only ATOM lines
New fields in Pdb for CAPS identifiers: passed all the way down to RIGraphs in getGraph()

Minor change of some error message

Fixed bug! actually it was not an abbot bug but our own, now calling AAinfo.class correctly (instead of System.getRunTime()....). With this abbot runs fine without any dirty fixes

Fixed to be able to run CMView under Abbot (because of abbot getResourceAsStream bug). The ABBOT_FIX flag is turned off.

checkConsurf: if the file is not found and the chain code given is "A", we check now also for results for NULL chain code 

Changed Alignment class so that both alignment and sequence indexing are starting at 1 (before alignment indices were starting at 0). Also mapping is now done through arrays not maps.
Changed all other classes using Alignment to accommodate this. 
NOTE: graph averaging hasn't been tested after the change

Removed PairwiseAlignmentGraphConverter, not needed anymore by CMView + was very buggy
Improved some comments/java docs

Interval: allow spaces in residue selection strings
RIGNbhood: added method getNeighbors

Latest scop and csa versions updated.

Fixing the fixed bug!
-if graph read is chain graph, then fullLength is set to the maximum observed residue serial
-else if graph is scop graph, then fullLength is set to the number of observed residues
NOTE:In the case of a wholechain scop graph, there might be inconsistensy between a graph I created with Pdb/RIGraph classes and stored in db and the corresponding graph i read from db. Specifically, the fullLength of graph I equals the length of the chain while the fullLength of graph i equals the number of observed residues.

Added getCommanSeparatedResSerials to RIGCommonNbhood

added some comments

Fixed bug in DSSP secondary structure assignment, first element was starting at index 0

New method to convert secondary structure

Now converting also secondary structure in pairwise alignment graph conversion

Fixed three compiler warnings for unused imports (DbRIGraph, Pdb) and an untyped iterator (CatalSiteSet)

createGraphDb:
-scop_graph table added
-!!!!residue serial fields (num, i_num, j_num) are changed to unsigned

!!!!ProtStructGraph:
-interSSE variable added
-getResidueSerial abstract method added
-restrictContactsBetweenSs method added

!!!!AIGraph:
-getResidueSerial method added

RIGraph:
-!!!!getResidueSerial method added
-scop graphs are written to db correctly
-interSSE variable taken into account for CR db field

DbRIgraph:
-changes made so to be able to read from db using a scop id
-interSSE and minSeqSep are set now in get_db_graph_info()
-!!!!FIXED BUG:fullLength in read_graph_from_db is not set to the size of the serials2nodes TreeMap instead of the maximum serial in serials2nodes. It was giving wrong result for scop graphs.

testDbRIGraph added:
It reads one graph from a source db based either 
-on pdbcode, chaincode and graph properties or 
-on scop id and graph properties or 
-on a graph id 
and it writes the graph to a destination db.

Pdb:
-unused scop regions are now removed in restrictToScopDomain
-unused residues are also removed from resser2pdbresser, pdbresser2resser, resser2allrsa, resser2scrsa, resser2consurfhsspscore, resser2consurfhsspcolor, catalSiteSet
-!!!!sequence is reset to scop sequence and fullLength to the length of the scop sequence

CatalSiteSet:
- removeCatalSiteRes(int resser) method added

CatalyticSite:
-remRes(int resser) method added

Scop:
-remove(ScopRegion e) method added

genDbGraph:Comments only added to demonstrate new functionality.
-comment added to show how to use restrictContactsBetweenSs
-comment added to show how to use restrictToScopDomain
-runDssp now is always run and not only when mode != "GRAPH". This has been changed since now the contact range might depend on the ss assignment (restrictContactsBetweenSs) and we want to ensure consistent results
-runNaccess moved to the bottom so it is always run last. In this way if restrictToScopDomain is used, we don't have to run naccess twice.

EC, CatalSiteSet: added typing to some iterators to avoid eclipse warning
Interval: added typing to Comparator to avoid eclipse warning

Changed implementation of Alignment: now based on indices rather than tags. Interface remains mostly the same

New functionality: now we can get sequences and reset tags based on the order the sequences appear on the file.

Got rid of check for equality of tags

Now also copying the CASP RR identifier fields in RIGraph.copy() and in PairwiseAlignmentGraphConverter.converGraph().

Extracted some constants: NO_PDB_CODE, NO_PDB_CHAIN_CODE, NO_CONTACT_TYPE, NO_CUTOFF

added main function and loading from Multi-model PDB files

Copied the aglappe-jung branch into trunk.

added new classes RIGEnsemble and FileTypeGuesser (used by Casp Consensus Contact prediction)

added missing class for RR file i/o

added reading from and writing to casp contact prediction (RR) files

New method calculateRadGyration

if the pdb file contains SEQRES records we do not allow the sequence length to be shorter than the full-length of the graph. otherwise a PdbfileFormatError is thrown in function read_data_from_file()

Adapted IO methods in NodesAndEdges to JUNG framework:
added 2 classes: 
- GraphIODb: does reading and writing a Graph<V,E> from/to database
- GraphIOGDLFile: does writing of a Graph<V,E> to GDL file for aiSee exporting

Now maxcluster ranking works in all cases. 
Fixed some bugs and tested for both rmsd and gdt

Moved max cluster methods from SimilarityGraph to MaxClusterRunner. Removed SimilarityGraph
Now MaxClusterRunner can do all modes of maxcluster: ranking, matrix, and pairwise
Changed default for DEFAULT_FORCE_CONSTANT in TinkerRunner to 10.0

new function: createSelectionString(TreeSet<Integer>)

Pdb objects now return strictly PdbLoadErrors (before, GetterErrors were thrown)

Fixed potentially dangerous bug: was constructing edges wrongly!: with i residue number as the weight!

Fixed bug: now assigning frequency 0 to case where countTotal=0 (no observations)

New feature: now Pdb objects can be restricted to SCOP domains
New member field sid in Pdb and ProtStructGraph if Pdb/ProtStructGraph is restricted to SCOP domain

added getter function for the cif-filename

Minor change: comments

Minor change: javadoc

Fixed bug: in removeVertex we were calling getResidueSerial instead of getAtomSerial

Now overriding removeEdge(E) to fix a bug in SparseGraph JUNG2 alpha2
Not implementing anymore SimpleGraph (was useless and it will be removed in coming JUNG2 releases)

moved GraphAverager to subpackage, added removeVertex function to RIGraph and AIGraph, added getEdgeFromSerials and containsEdgeIJ functions to ProtStructGraph

Now also reads secondary structure

Minor change : fixed comments

Fixed bug: getCommaSeparatedResSerials was failing when there was 0 neighbors

New class NbhoodBackgrnd representing a set of neighbourhoods loaded from a set of graphs of a non redundant pdb structures set.
Rewrote RIGNbhood so that the TreeMap now contains also the central residue. Added methods equals() and match().
Adapted RIGraph.getNbhood to the new constructor in RIGNbhood

class *Pdb:
- implemented new exception handling scheme for function Getter.get() -> getChains() and getModels() throw all possible exceptions encapsulated in GetterError exceptions

Pdb.java:
-calc_atom_dist_matrix method added
-calc_dist_matrix works also for multi atom contact types by choosing the min distance

Now not using ORDER BY in nodes and edges queries, was useless and slowing down queries
Now reading correctly empty graphs, just keeping it empty without throwing exception

Now Pdb constructors don't load data but rather intialise the pdbCode, loading occurs upon call of load(pdbChainCode,modelSerial)
New methods getChains() and getModels() in all Pdb classes
New Exception PdbLoadError
New tester for getChains and getModels: testGetChains
Changed all calls to Pdb construction accordingly (includes changing excpetions)

AAInfo
-isValidContactType, isValidSingleAtomContactType, isValidMultipleAtomContactType takes into account the directionality, the overlapping atoms and the "+"
-isOverlapping method has been added

AIGEdge
-copy method added

AIGNode
-equals and copy methods has been added

AIGraph
-getRIGraph now takes argument directed and directionality doesn't depend anymore on isCrossed
-distances are now set for RIGNodes in getRIGraph
-for undirected, crossed graphs we take care not to add parallel RIGedges in getRIGraph
-we set now the secondaryStructure object of RIGraph in getRIGraph
-addGraph has been added

Box
-we throw away wrong check in getDistancesWithinBox - read the comment

DbRIGraph
-constructors now take argument directed and directionality doesn't depend anymore on isCrossed
-distances are read also from db and set for RIGEdges

Pdb
-get_graph now takes as argument directed and directionality doesn't depend anymore on isCrossed
-there is check for forbidden cts in get_graph
-cts with + are taken into account in get_graph

RIGEdge
-setAtomWeight and setDistance methods have been added

RIGNode
-equals method has been added

RIGraph
-getObsLength has been deleted
-we take care now of the new cts (+,undirected crossed), of null secondary structure case and of distances in write_graph_to_db and write_graph_to_db_fast

genDbGraph
-argument for directionality is now added

Reverted to unobserved RIGNodes not present in the RIGraph

Changed behaviour of observed/unobserved nodes in RIGraph: now unobserved RIGnodes are also in the RIGraph, tagged with observed=false.
Thus getObsLength() in ProtStructGraph (which by the way was not working for AIGraphs) is now only in RIGraph and adapted to the new way of tagging unobserved nodes.
New method hasSequence() in ProtStrucGraph and Pdb: it is important to know that from now only if hasSequence()==true we put the unobserved nodes in RIGraph, otherwise we don't put them (as before)

Added more comments

Fixed bug: was missing last sequence element when adding gap nodes

Fixed bug in PairwiseAlignmentGraphConverter: was not adding the nodes for gaps, also some other minor bugs fixed in copying of objects
Fixed bug in getAllCommonNbhSizes, 1) now returning a HashMap on 
Pair<Integer>, not Pair<RIGNode> 2) was filling cells when size=0
Added method getIncidentNodes to IntPairSet as a temporary fix

Now PairWiseAlignmentConverter and PairwiseAlignmentGraphConverter work in the new JUNG framework.
New class IntPairComparator
Now sorting output in RIGraph.write_graph_to_file

Added last i/o method from old Graph (write to file in Ioannis network format)
Using now the more correct degree, indegree, outdegree methods instead of count neighbor methods
Fixed the write_graph_to_file method name in executables in default package

New methods in RIGraph copy() and compare(other)
Basic testing of copy introduced on testJUNGframework

Now selection string stuff working: transfered methods from old NodeSet to Interval as static methods.

Now graph generation in Pdb also works when we pass a crossed contact type with overlapping atom sets, e.g. ALL/BB (changes were: added a couple of new conditions in Box and in Pdb's AIGraph getGraph(ct,cutoff))
Loads of better comments

First implementation of the new graph framework using JUNG2. Still a few classes to fix, but all basic functionality is there.
NOTE: at the moment ProtStructGraph is a SparseGraph which doesn't guarantee that the graph has no parallel or loop edges (I thought that would be solved by using the SimpleGraph interface but it doesn't, that's only a marker interface)

Re-branching for JUNG2 development

Minor changes. Got rid of some warnings

Changes in Pdb
-checkConsurfHssp: in the mistakes calculated I add now at the end the difference between the number of residues and the number of consurf scores

-writeAtomLines has now a boolean argument called pdbCompatible: if pdbCompatible is true, then only the first character of chainCode and not the 'whole' chainCode is exported to the file.
-another WriteAtomLines has also been added without this extra argument for backwards compatibility: default value for pdbCompatible is false
-dump2pdbfile has also been changed in similar way
-runDssp and runNaccess now call the above methods with true pdbCompatible

Changes in genDbGraph
-1 summary output line has been added to provide info about csa, naccess, scop, ec, dssp, ...
-DSSP is run with 3state

minor change

minor change

contactTypes.dat 
-SC_CAGLY contact type added

SecStrucElement.java
-getSheetSerial method is now public

Pdb.java
-writeToDbFast added

Graph.java
-write_graph_to_db_fast
-expBB,CW,CT,CR,w,d are now taken into account in a correct way
-node_id, sssid, sheetSerial and turn are now also filled in

DbGraph.java
-expBB,CW,CT,CR,w,d are now taken into account in a correct way
-directionality is also taken into account now when reading edges

genDbGraph.java
-sequence separation added as option

createGraphDb.java added

Reverting to last revision, changes made in 405 were wrong

Alignment and PairwiseAlignmentConverter now using Interval, IntervalSet and TreeSet<Integer> to replace Edge, EdgeSet and NodeSet
New class IntervalSet

Not asserting anymore pdbChainCode to be uppercase. pdb chain codes are case sensitive!

OXT is taken into account now also for BB contact type

Forgot CiffilePdb!: not forcing upper casing of chainPdbCode. pdb chain codes are case sensitive!!!

Not forcing upper casing of chainPdbCode. pdb chain codes are case sensitive!!!

-pdb_seq_num is used instead of auth_seq_num in get_ressers_mappings. Now resser2pdbresser contains all residues even the unobserved ones. This has been changed in PdbasePdb and CiffilePdb classes.

-pdb_strand_id not needed in the query in read_seq in PdbasePdb class. Asym_id and pdb_strand_id have 1-1 mapping in case sensitive mode.

-comments have been added about the entity_id-asym_id-pdb_strand_id 1-1-1 mapping and some filters in queries that are no longer necessary.

added class PairwiseSequenceAlignment

Pdb code forced to lowercase in CiffilePdb

New method getMotifReducedAlphabet to get neighbourhood strings in Michael's reduced style (must pass a Graph with weights which indicate BB or SC domination)

started implementing a wrapper for the Maxcluster application (some related code is now in SimilarityGraph but should be moved to MaxClusterRunner in the future)

New method setSequence

deleted the print part in the checkCSA method in pdb class

NEW TYPES
---------
- FileFormatError
- FastaFileFormatError
- PirFileFormatError
- AlignmentConstructionError

MODIFICATIONS
-------------
class Alignment:
- introduced new exceptions to be throw whenever file format error have been detected or any other construction errors occur
- changed accession rules of function getGapCharacter from "public" to "public static"
- new functions:
 * writeFasta
 
class CiffileFormatError:
- changed super class from Exception to the new FileFormatError class

class GraphFileFormatError:
- s. class CiffileFormatError

class PairwiseAlignmentConverter:
- added exception handling features due to class Alignment

class PairwiseAlignmentGraphConverter:
- s. class PairwiseAlignmentConverter

class PdbFileFormatError:
- s. class CifffileFormatError

hasCSA and getCSA methods added in Pdb.java

inSameSheet, isOther and getSheetSerial methods added in SecStrucElement

Methods added for:
-parsing CSA
-parsing EC
-parsing CONSURF-HSSP
-running & parsing Naccess
-parsing SCOP
-writing residues' info in a db table

Modifications have been also made in DSSP parsing:
-sheet serial is taken into account
-flexibility in reduced state preferred

added a new version of getDiffDistMap(...) that take a sequence alignment as an input to allow for residue mappings between two pdb-models of different size

redesign of TinkerRunner: added 2 reconstruct methods, one which generates normal tinker output for a given number of models, and one fully automatic one which simply returns one pdb object;
updated reconstruct.java to make use of the new TinkerRunner method;
updated GraphAverager: now assigning edgetype and cutoff to consensus graph

finally implemented the setCutoff method ;)

fixed nasty bug in constructor Graph(sequence) (residue indexing was wrong)

added method setCutoff() and some javadocs

Fixed bug: getContacts was not copying weights
Got rid of useless copyKeepingNodes(), made main Graph constructor public. Some new setters and getters

New method getEdge in EdgeSet, necessary since we introduced weights in the Edge class. To retrieve them we first need to retrieve the actual Edge object

Method writeUndirUnweightGraphToNetworkFiles added in Graph class

renamed precision to accuracy (in accordance with Casp)

Now weights are kept in the Edge objects rather than separately in a member variable in Graph

Graph,PredEval: added fields coverage and title to predEval (and evalPrediction)
GraphAverager: added method to return average graph (with weighted edges)
PdbfilePdb: fixed bug when reading from Casp prediction files

Fixed bug: now does correctly quoting for any of the 3 cases: "", '' or ;;. Does also correctly multiline parsing of ;; quoted fields. Passed testing on cullpdb 90 (more than 9000 pdb entries)

Fixed bug: wasn't tokenising well when first field in line was quoted

FIXED BUG: now doesn't fail with records that are delimited with \n; ;\n
Method tokeniseFields is now completely rewritten: is what does all the magic of parsing all the oddities of the mmcif format
Using RandomAccessFile to open the file only once and then seek to the positions we need to scan at each point. Might be slower due to the RandomAccessFile that does no buffering. Also maybe because the new tokenisation is not very optimal
Now parseCifFile does the whole parsing calling also the submethods instead of calling them in the constructor

Changed constructor with all parameters to protected, now shouldn't be used at all outside of package

Major redesign of NodeSet, now contains objects of new class Node. Nodes are essentially integer which can have additional properties. Added equals methods to NodeSet, EdgeSet, NodesAndEdges. Moved some methods from SimilarityGraph to NodesAndEdges.

Fixed documentation

Added two new constructors for Graph (sequence only and no parameters), changed interface of GraphAverager (pass target sequence instead of target graph), extracted constant DEFAULT_WEIGHT in Edge.

extracted constant NULL_CHAIN_CODE from ...Pdb classes, added copy() methods to NodeSet and EdgeSet, added some functionality to NodesAndEdges, new class SimilarityGraph

Now ConstraintsMaker works also for multi atom contact types (BB and SC) and crosses (either BB/SC or any cross to a single atom contact type)
New file aapairsBounds.dat with values for aminoacid side chains distance lower and upper bounds taken from statistics of a set of pdb representatives
New in AAinfo: aapairs2bounds and changed getUpperBoundDistance, getUpperBoundDistance to accomodate aapairs2bounds
PRMInfo: now for amber "OXT" gets transformed into "O"

New class NodesAndEdges (simple graph class), some added javadocs

added constructor for trivial alignment, renamed getSequence to getAlignedSequence

Better handling of graph averaging for different thresholds
Refactored predict to doAveraging, not returning anymore a Grap but rather modifying the reference

added a weight field to Edge class (set using alternative constructor, maintaining backwards compatibility)

Added GraphAverager class and PredEval class. GraphAverager copied from HomologyPredictor (in CMPredict): doing the averaging correctly now
Added method evaluatePrediction to Graph
Some refactoring in Alignment and better javadocs

First commit of ConstraintsMaker (and PRMInfo needed for it)
Does correct atom serial mapping when using amber prm files, except for some atoms of ASP, ARG, GLU, PHE, TYR (where there are some nomenclature differences between amber and pdb)
Does not work yet with multi atom contact types, should work fine for all single atom contact types and for crossed single atom contact types
The pdb/xyz atom mapping using coordinates (as in python) is commented out because it doesn't work due to rounding differences between java and fortran
Thus the PRMInfo class was needed, at the moment can read only amber prm files
Some new methods and better javadoc in Pdb.
Some new methods in AAinfo

Removed class AA and replace it by AAinfo, which reads contact types from separate file contactTypes.dat
New class ContactType which contains atoms for each contact type and residue. A static object for each contact type is loaded into AAinfo upon reading the contactTypes.dat file
Changed all references accordingly

added parsing of residue selection strings

2 new methods: getAllAtomSerials, getAtomCoord

Moved tinker stuff to its own package

Initial commit of TinkerRunner and exception class to go with it TinkerError

added calculation of difference distance map

added constructors for loading from online pdb

added some javadoc comments

Fixed some comments

Now parsing each element in different methods (re-opening the file). Parsing first pdbx_poly_seq_scheme so we get the chainCode that we can use for reading the rest
Now taking care of cases where struct_sheet_range is not a loop element
In tokeniseFields now also unquoting double-quoted strings
Tested on a set of 12000 entries

Checking number of fields per line in loop elements and throwing exception if count is not correect
Doing tokenisation of lines through new function that takes care of possible quoted string with spaces
New exception CiffileFormatError
Checking 1st line of cif file has correct format: data_1xxx, if not throwing exception

Fixed buf: sometimes struct_conf can be non-loop elements, now also taking care of that particular case

Bug with '?' in auth_seq_num was not really fixed. Now should be fine: behaviour is the same as PdbasePdb

Fixed bug: needed to read alt locs in advance in another scan of the file because the order of the elements in the cif file is not guaranteed. As read of atom_site needs of alt locs, we need to do first the parsing of atom_sites_alt

Fixed bugs:
- was reading HETATM lines as well as ATOM in atom_site
- auth_seq_num with '?' not taken now when populating the pdbresser2resser map (same behaviour as in PdbasePdb)
- now using chainCodeStr and auth_asym_id to identify chains in pdbx_poly_seq_scheme, struct_conf and struct_sheet_range. atom_site is not guaranteed to appear in file before all the others so we can't rely on having read a chainCode (asym_id) when parsing the other elements

Now taking indices for fields from parsed field names. Still only minimal testing

First implementation of mmCIF file parser. Tested minimally.

Got rid of useless (and potential bug) chainCode=getChainCode (self reference!)

changes to calcGridDensity, now also counts number of neighbouring non-empty grid cells

changed constant for unknown pdb code from "Unknown" to XXXX

Better checks for sequence equality in compare

Now also filling chain_pdb_code in write_graph_to_db

Fixed bug in  compare, wasn't assigning new weights

Introduced weights in Graph objects: we can read them from file or db and write them into file or db. On calling pdb.get_graph the weighs are taken as number of atom-edges per residue-edge (similar to Ioannis weights in chain_graph tables)

Wasn't deep copying secondary structure when passing it to Graph

Made get_graph a bit less memory hungry, by using a float[][] instead of double[][]
Not printing warnings about missing atoms anymore

fixed little bug in NodeNbh

changed internal implementation of runDssp method (not using getFullLength anymore)

Fixed bug in Graph constructor: was not getting correct fullLength (related to last bug fix)

Fixed bug with getFullLength when reading from pdb file
Now if SEQRES was missing then fullLength is maximum observed residue number instead of total number of observed residue numbers
New member variables obsLength and fullLength

Added secondary structure to write_graph_to_db
New method getNumSecStructureElements in SecondaryStructure

Fixed bug: wasn't taken into account OXT atoms in calculating the ALL graph

Added new SecondaryStructure object to Pdb and Graph

Fixed potential bug in regex for ATOM lines

Was still not fixed!, forgot to insert X for non standard aa

Fixed reading of SEQRES, was putting nulls for non-standard aas

Implemented write_graph_to_db
New member variable model in Graph, thus changed constructor
New methods to get degree of a node in Graph
New methos in NodeNbh to write other kinds of neighbourhood strings

completed first version of dssp secondary structure assignment, can be switched on/off by status variable in Start, some other little fixes

Now reading sequence from SEQRES lines. This fixes bug in cmview window was not showing whole contact map because of size of contact map was set to number of observed residues size

fixed bugs in ATOM line export (atom names were wrongly aligned & added three more columns after coordinates), added method to run DSSP (still incomplete)

Added getNumAtoms

PdbfilePdb now also accepts Casp prediction files ('PFRMAT TS' instead of 'HEADER')

Made constructor public

added application to count grid density, suppressed warning for secondary structure reassignment

New methods containsContact and resetContacts

Added method hasTag

Made regex a constant and added | and \\-

Doing now everythin with maps (using seqTags) rather than arrays. Cleaned up, refactored, using more consistent naming

Added reverse map (from sequences serials to alignment serials) and did a bit of refactoring

Initial commit of Alignmnent, a class representing a multiple sequence alignment

Got rid of doneRes ArrayList in getRanks, was redundant

Now dealing correctly with ties.
Implemented getResiduesSortedByRank
Changed dist to be a TreeMap

Fixed bug in delEdge. Wasn't taking care of if contact was given as i>j in undirected case. Also recount of contacts and set of modified to true was wrong

Fixed bug in addEdge method. In case of undirected we need to check that the contact is inserted in the right order j>i

Optimised slightly the compareTo method in Edge

Now rmsd calculation works for 2 Pdb objects where there are missing residues (unobserved) or missing atoms
New Exception ConformationNotSameSizeError
Still not printing out individual distances between 2 conformations (altghough it is possible)
Can't specify yet a restricted range of residues to calculate rmsd only with them

Fixed bug. Was not doing correctly reflection along smallest principal axis in is_reflection case

Implemented rmsd calculation (based on Jama matrix package)
rmsd method still not taking care of unobserved or missing atoms. Also not returning individual distances between optimally superimposed structures (although it can calculate them in commented out code)

Made ContactList a TreeSet instead of an ArrayList, should improve performance (log(n) rather than linear)
REFACTORING: 
ContactList -> EdgeSet
Contact -> Edge

Added method getCopy

Now checking in addEdge that we don't duplicate edges
New methods: get2ndshell, copyKeepingNodes

added class NodeSet

Now calculate_dist_matrix returns the matrix with residue serials as indices (rather than atom serials)
Now AA.isValidSingleCT and AA.isValidMultiCT check also for crossed cases

Now can do crossed contacts also with geometric hashing
Renamed getGraphGeometricHashing to get_graph (so interface remains as it was before we introduced the geometric hashing)

First geometric hashing implementation for calculating graph (with new class Box). Only does non-crossed cts
Geometrich hashing graph calculation kept in a separate method than normal get_graph. Still either of the methods can be used.
Using javax.vectormath objects for the vectors (distance method also from that)
Using TreeMap for get_coords_for_ct

Now calculate distance matrix uses HashMap instead of TreeMap
Needed to change the hashCode function in Contact to make it faster

Added cross contacts to isValidCT

New secstruct2resinterval TreeMap to store secondary structure elements as a map of ss string ids to intervals
New class Interval

Now reading also secondary structure (in a map of residue serials to secondary structure identifiers)

now casting pdb codes to lower and pdb chain codes to upper in all constructors

Fixed bug: ATOM lines printing was locale dependant, thus was printing ',' for decimal separators in some locales

some clean up of exceptions in Pdb and Graph classes:
merged PdbAcCodeNotFoundError and MsdAcCodeNotFoundError into PdbCodeNotFoundError,
created new PdbChainCodeNotFound, got rid of stderr output, throwing exceptions with informative messages instead

Fixed bug in regex introduced in last revision

Fixed graph file format checking, was not working properly

Now throwing also SQLExceptions for individual queries, not just for connections

Throwing exception also in case graph file not in right format at all (no #AGLAPPE in 1st line)
Fixed bug: was not reading correctly #PDB CHAIN CODE

FIXED BUGS:
- directed was not set when reading from cm file
- pdbChainCode not set when reading from db given pdbCode
- filling pdbresser2resser and resser2pdbresser hashmaps also in reading from pdb file
- using "A" as chainCode when reading from pdb file
- some error handling for file formats: new exception classes GraphFileFormatError and PdbfileFormatError

Restructured construction of Pdb and Graph objects: now subclasses for each case
Cleaned up and made consistent database connections
Now can also pass a MySQLConnection in all cases (as well as having default values for a default connection)
PdbaseInfo and MsdsdInfo classes removed: now merged into PdbasePdb and MsdsdPdb respectively
Updated following this changes testPdb and compareCMs

Major refactoring:
 accode -> pdbCode
 chaincode -> pdbChainCode
 chain -> chainCode
Major change of signatures (publics to privates)
Implemented getters for needed member variables
Major clean up

Using apache xmlrpc version 3.0 in PyMolServerOutputStream

MySQLConnection now throwing SQLException on connect
Many files changed following this: all calling classes now re-throwing or catching the SQLException

New isValidCT methods in AA
New method get_resser_from_atomser in Pdb

Implemented getRanks()

AA class is now db independent:
Hard coded getaas2atoms, not getting it anymore from db.

Added toString and getMotifFullGaps missing in EdgeNbh

Made contactTypes a String[]

Added contactTypes method and made private methods private

Moved motif static methods to Background class. They didn't belong here

Made contacts public

New method getAllEdgeNbhSizes

NEW FEATURE: mapping of pdb residue serials and internal serials in the Pdb object

Added basic compare method to the Graph class and test class compareCMs

FIXED BUG: The returned regex in motif2regex was wrong, needed a ".*" at the beginning

Implemented hashCode method in Contact and GenericContact. Was necessary to use these classes as keys in Maps
Fixed bug in equals method in GenericContact
Changed middle character from "\t" to " " in toString

New class GenericContact. To store a generic contact, i.e. a pair of residue types

Fixed bug: calculateEntropy was wrong when a pi was 0

FIXED BUG: the switch statement was missing the breaks
This was causing strange behaviour in counts NodeBackground getSize methods and also was making them being too slow

New class NbhProbDistribution
In NodeNbh implemented a few static classes to convert between different motifs and to match motifs using differing criteria

FIXED BUG: fixed restrictContactsToMax/MinRange. Couldn't delete contacts at the same time as looping in them
NEW CLASSES NodeNbh and EdgeNbh, now used in getNodeNbh and getEdgeNbh
New method getResidueType

Fixed bug in getMaxRange method, was not assigning to the max variable
Added toString method to Contact

FIXED BUG in initialisation of fullLength when reading from file. In case of no sequence and nodes provided, was not getting correctly the maximum value for contacts. Now using the method getMaxContact from new class ContactList
NEW FUNCTIONALITY in Graph:
-New member variable modified
-New methods addEdge, delEdge, restrictContactsToMaxRange, restrictContactsToMinRange, getContacts, getNodes, copy
-Improved slightly the implementation of getEdgeNbh 
FIXED BUG in initialisation of fullLenght when reading from file. In case of no sequence and nodes provided, was not getting correctly the maximum value for contacts. Now using the method getMaxContact from ContactList
New class ContactList
NEW FUNCTIONALITY in Graph:
-New member variable modified
-New methods addEdge, delEdge, restrictContactsToMaxRange, restrictContactsToMinRange, getContacts, getNodes, copy
-Improved slightly the implementation of getEdgeNbh 
New method getRange in Contact

Got rid of some unnecessary warnings

New functionality: added methods to get node/edge neighbours

made Graph.directed public

Moved ContactMap and EdgeState out of this repository. 
Extra functionality of ContactMap now in Graph:
- new member variables fullLength, obsLength, numContacts. New method getIntMatrix
- removed getCM method

FIXED BUG in get_db_graph_info, was not dealing well with the case of chaincode being NULL in the database

FIXED BUG: in getCM when directed was passing a reference to the contacts object and not deep copying it!

Now passing also chaincode from Pdb object to Graph object in get_graph

Fixed bug: wasn't initialising accode, chain and chaincode when reading to file. Now all set to "" when not present in file

Recoded chain reading in read_pdb_data_from_file to make it clearer, no change in functionality

NEW FUNCTIONALITY: Reading from pdb file given chain code or model serial fully implemented
Changed the chain that is dumped in dump2pdbfile directly to the Pdb.chain field i.e. internal chain identifier
Loads of new comments, including java doc for all constructors

FIXED 2 BUGS. NEW FEATURE reading graph from db given graph_id
Bugs:
- nodes TreeMap in read_graph_from_file contained one letter res codes, now three letter codes
- in getCM was passing residues as a TreeMap of nums to three letter residue types, now passing one letter residue types
New constructor for getting graph from db given graph_id, new method to go with that get_db_graph_info

Made cutoff and ct public

Made some fields public: sequence, accode, chaincode, model

NEW FEATURE: Read and write graphs from file fully implemented.

Refactored read_contacts_from_file to read_graph_from_file. Now can read other fields as well as contacts. Also gets nodes if sequence present
New method write_graph_to_file

FIXED BUG: last commit didn't fix bug correctly:
Changed the ContactMap constructor to put into residueNums and residueTypes nums from 1 to maximum of residues.keySet()

Made public accode and chain fields in Graph

FIXED BUG: getCM wasn't working well when reading from db and there were unobserved residues

Now setting the sequence to blank when reading from db. Also when reading from file (still nodes reading is not implemented)
In ContactMap constructor different handling when sequence is blank

Changed input in constructor for reading from db: chain->chaincode
Now getgraphid initialises too the internal chain code (pchain_code from graph db)

Made Pdb.acccode public

Changed default accode and chaincode to "Unknown" when reading from pdb file

FIXED BUG: wasn't setting the chain field when reading from msdsd

Made chain field in PDB class public

FIXED BUG: pdb data load from msdsd wasn't working for NULL chaincodes

NEW FUNCTIONALITY: reading of graph from db is fully implemented for all cases

New chain member variable in the Info classes, read in get_asym_id (Pdbase) and in get_chain_id (Msdsd)
Reading chain also in Pdb in read_pdb_data_from_file
Not reading chain anymore in read_atomData of PdbaseInfo
Added oneletter2threeletter and getoneletter2threeletter to AA class
Changes in Graph:
- added db static vars and getUserName method
- new member variables graphid and sm_id
- new method read_graph_from_db to read contacts, nodes (and sequence from nodes) from db
- new method getgraphid
New Exception class GraphIdNotFoundError

Implemented basic reading contacts from file

New functionality: now can create a Pdb object passing a model serial for NMR structures. Not implemented when reading from pdb files!

2 new convenience methods: getBoolMatrix and getIntMatrix
Got rid of method to print constraint equations, was useless

NEW FEATURE: reading pdb data from msdsd
New classes MsdsdInfo and its exception classes
FIXED BUG in dump2pdbfile in Pdb: was not sorting the atom lines

NEW FUNCTIONALITY: can retrieve now a ContactMap object from a Graph object
New method getCM in Graph. New member variables: sequence, accode, chain, nodes, directed
Passing more data from Pdb to Graph in get_graph
New constructor in ContactMap

ContactMap and EdgeState classes moved here from CMPredict repository

Implemented read of pdb data from pdb file.
FIXED BUG in dumping of pdb file. Was missing end of line in atom lines
Added more tests in testPdb

MYSQLUSER now taken from unix name with new getUserName method

Added error handling for PdbaseInfo: 2 new exception classes

Initial commit of the proteinstructure package. Contains classes to use pdb protein structures