Toxtree-v1.40 (build date 24 Mar 2008):
- improved descriptor calculations (I(An),I(BiBr));
- updated QSAR6, QSAR8, QSAR13;
- new decision tree layout based on prefuse library - can be exported as .png, .bmp, .jpg or .gif;
- new decision tree export options - .html, .pdf or .txt;
- user preference options (persistent across application instances);
- updated user manual;
- updated README;
Toxtree-v1.36 (build date 17 Feb 2008):
- improved partial molar refractivity, Sterimol and Idist descriptor calculation;
- updated README;
Toxtree-v1.35 (build date 3 Feb 2008):
- updated the structural alerts for predicting carcinogenicity and mutagenicity;
- updated descriptor computations;
- updated README;
Toxtree-v1.34 (build date 20 Jan 2008):
- updated CDK library;
- updated the structural alerts for predicting carcinogenicity and mutagenicity;
- updated descriptor computations;
- updated README;
Toxtree-v1.33 (build date 6 Jan 2008):
- updated to CDK 1.1 release;
- updated the structural alerts for predicting carcinogenicity and mutagenicity to take into account overlaps;
- updated README;
Toxtree-v1.32 (build date 26 Sep 2007):
- support for assigning multiple labels;
- improved implementation of structural alerts for the decision tree for predicting carcinogenicity and mutagenicity;
- MOPAC call bug fix;
- updated README;
Toxtree-v1.31 (build date 21 Aug 2007):
- initial implementation of a decision tree for predicting carcinogenicity and mutagenicity - 35 structural alerts and 3 linear discriminant rules, making use of 11 descriptors;
- new descriptors implemented:
- Partial Molar Refractivity of substituents, according to a lookup table
- Sterimol L & B5 descriptors, according to a lookup table
- Molar Refractivity (uses Joelib implementation of "S. A. Wildman and G. M. Crippen, Prediction of Physicochemical Parameters by Atomic Contributions, J. Chem. Inf. Comput. Sci., 39, 868-873, 1999")"
- Descriptors for structure presence of aniline, NO2 & bridged biphenyls
- EHOMO, ELUMO and other electronic descriptors calculated by launching MOPAC 7.1 (http://openmopac.net) with fixed options (AM1)
- added substituents.sdf file (substituents lookup table);
- added helper directory, containing MOPAC_7.1.exe;
- updated README;
Toxtree-v1.30 (build date 19 Jun 2007):
- updated to CDK 1.0 release;
- updated to jre-1_5_0_12;
- prototype implementation of a decision tree for predicting carcinogenicity and mutagenicity;
- updated README;
Toxtree-v1.20 (build date 11 Mar 2007):
- Toxtree 1.20 at SourceForge
- added global options for skin irritation rules (in addition to rule specific options);
- added "Silent" option for skin irritation property rules (the answer of the property rule will always be "No" if a property is missing);
- improved implementation of Verhaar scheme Class 4;
- fixed error when writing SMARTS rule to .tree file;
- updated installation manual;
- updated user manual;
- updated README;
Toxtree-v1.12 (build date 18 Jan 2007):
- added new rule type based on SMARTS;
- redesigned decision tree editor;
- added Verhaar scheme plugin (Toxicity mode of action prediction);
- added BfR rules plugin (skin irritation/corrosion prediction);
- updated CDK library;
- updated README.
Toxtree-v1.00 (build date 23 Oct 2005):
- added complete application documentation, including installation manual, user manual, updated README, documentation of the source code;
- improvements of the decision tree editor.
Toxtree-v0.05 (build date 14 Oct 2005):
- user defined decision tree concept introduced;
- decision trees can be saved to and loaded from a file (a custom ".tree" file type is used);
- added graphical user interface for building decision trees;
- the structure diagram editor (JChemPaint) is now launched as a nonmodal dialog. This resolves the issue of some JChemPaint windows being inaccessible if the editor is launched as a modal dialog;
- after structure is edited in JChemPaint, SMILES is created and other properties are retained;
- new command line options are introduced:
java -jar Toxtree.jar -f input file for normal open
java -jar Toxtree.jar -b finput file for batch processing
- improved recognition of carbohydrates;
- several nucleotides added to bodymol.sdf according to an expert advice;
- better recognition of some ionic groups (note that [Ca++] is recognised only as [Ca+]. For correct representation use [Ca+2] in SMILES);
- batch processing can not be started anymore when input and output files are the same;
- Q11: when a heterocyclic ring substituents are analysed, only heteroatoms in the ring analysed are disregarded (as specified in the Cramer et all, 1976) but not the heteroatoms in other rings. This behaviour is the one of v0.03. In the v0.04 release the behaviour was changed to disregard all heteroatoms, which doesn't follow the specification of Q11 to analyze each ring separately, with the remainder of the structure as substituents of that heteroring;
- added the application source code in the installers;
Toxtree-v0.04 (build date 04 Oct 2005):
- Q5: better heuristics for carbohydrates;
- Q6: return "yes" only if the a)- and b)-substituents are present simultaneously;
- Q9: if it is a lactone from this point on treat the structure as if it were the hydroxy acid;
- Q11: ketone changed to R-C(=O)-R where R is not (O or H);
- checked the search path for bodymol.sdf and foodmol.sdf, when the application is run from the installer;
- checked that compounds are assigned always the highest (most hazardous) class of their constituents/hydrolysis products;
- optimized searching in foodmol.sdf and bodymol.sdf;
- updated foodmol.sdf and bodymol.sdf according to an expert advice;
- added Back/Forward buttons for easier navigation in the history of the SMILES drop down list;
- made the SMILES entry panel inaccessible for user input, when there is an open file. The name of the file is displayed instead of the SMILES entry panel. If changes are made in the file contents in memory, an asterisk (*) is displayed after the file name. The asterisk is removed after saving the file. Access to the SMILES entry panel is restored after File/New.
- added keyboard shortcut (hotkey CTRL+SHIFT+T) for application launching;
- added Start Menu tool tips;
- minor enhancements of the installers;
- updated README.
Toxtree-v0.03 (build date 30 Sep 2005):
- included a 2D molecular structure editor based on JChemPaint. It could be accessed via the Compound/Edit menu;
- processes correctly SDF, CSV and TXT files with empty molecules;
- fixed incorrect perception of a specific functional groups configuration;
- bodymol.sdf and foodmol.sdf updated according to an expert advice;
- minor enhancements of the user interface;
- updated README.
Toxtree-v0.02 (build date 25 Sep 2005):
- Q1 and Q22: implemented to look up for "normal constituent of the body" and "common component of food" in bodymol.sdf and foodmol.sdf respectively. Only a very limited number of molecules are listed currently in these files. The user can replace these files with his own files, provided the names are the same. The files are expected to be in the same directory as Toxtree.jar (the executable file of the Toxtree application). If the files are not found there, the corresponding rules behave as "NOT IMPLEMENTED" and answer "NO" for every compound;
- Q4: treat the compound as free acid, amine, unsulphonated or unsulphated compound except for the purposes of Q24 and Q33 (as specified in Q4 of Cramer rules);
- Q15, Q17, Q29, Q30, Q31: hydrolysis products (if any) are processed separately, as required by these rules. If the products are assigned different classes, the entire compound will be assigned the most hazardous class (i.e. - if one product is of class I and another of class III, the entire compound is considered of class III);
- ionic representation of compounds is processed correctly now;
- checked that the following input file types are processed correctly (File/Open):
CML CSV - comma delimited, has a header; HIN ICHI INCHI MDL MOL MDL SDF MOL2 PDB SMI - each line starts with a SMILES, optional name delimited by single space, no header); TXT - tab delimited; XYZ - input/output of CSV and TXT: a column with "SMILES" heading is mandatory in order the structure to be read. All other fields are optional, will be read as molecule properties and displayed as such;
- all the properties from SDF files are read;
- output (File/Save): molecules can be written to SDF, CSV or TXT files, together with their Cramer classification data (class & path);
- decision tree results are written as SDF properties or CSV/TXT columns. The class assignment is stored in a property "toxTree.tree.cramer.CramerRules". The tree path is stored in a property "toxTree.tree.cramer.CramerTreeResult";
- added a batch processing facility. It allows faster processing without loading simultaneously all the compounds into memory. Please, use batch processing if the file has more than thousand compounds. A batch could be paused, continued after pause or cancelled. Batch configuration (i.e. input/output file and processed records) could be saved in a file, loaded and continued in a later moment, even after the application has been closed and started again later. Batch configuration is saved in a temporary file by default. This file is deleted if the batch completes successfully. If the batch configuration is saved into a user defined file, it will persist even after the batch completes;
- moved the SMILES entry panel from the left to the top of the application window;
- history of entered SMILES' is kept in a drop down list;
- the left panel now displays all the properties defined for a molecule (not only CAS, SMILES and NAME as in the previous version). The properties are read from a file (e.g. CSV columns, SDF fields) or set when decision tree estimation is performed. When the compound is created from SMILES (using the SMILES entry panel at the top) the property SMILES is set as well as a remark "Created from SMILES";
- compounds, consisting of more than one disconnected moiety, are now displayed;
- removed the "Explain" button on the right panel, since the result is automatically explained on estimation. A check box "Verbose explanation" is added instead. It affects the way results are explained;
- since there can be more than one hydrolysis product, the results for each product are displayed in brackets ();
- the application can be configured to output a lot of information on its behaviour. This could be useful for debugging as well as for detailed study of decision tree results. In order to start the application and output debugging messages to the console, type on a command line the following command:
java -DtoxTree.debugging=true -jar toxTree.jarTo store log in a file use:
java -DtoxTree.debugging=true -jar toxTree.jar > toxTree.log - minor enhancements of the installer;
- updated README.
Toxtree-v0.01 (build date 24 Aug 2005):
- first public beta-release of the Toxtree application.