ACD/Solubility DB
What's New
Among the key new features of ACD/Solubility DB presented below, are:
- A model for prediction of logP
- Enhancements to the pKa prediction model
- ACD/Quick Report
New Model for LogP Prediction
A generation model for prediction of logP is introduced which combines data for ~12,000 compounds from ACD/Labs' internal training set with new experimental data from >13,000 pharmaceutical lead compounds. Changes to the logP model impact the solubility module since calculation of intrinsic solubility requires logP.
NOTE: Both version 10 and 11 models for solubility will be available for use in training and prediction.
Enhancements to pKa Prediction Model
Version 11 of Solubility DB is improved due to the enhancements to the pKa model which helps describe the pH dependence of this property. The enhanced pKa model boasts the addition of >2,000 pharmaceutical lead compounds to the internal database, and over 600 new and revised Hammett equations.
ACD/Quick Report
A PhysChem interface has been developed for version 11 that offers a snapshot of the molecular property profile of individual molecules through customizable, color-coded histograms. Quick Report includes all descriptors available to the user from any ACD/Labs commercial PhysChem product.
Changes to training:
- The availability of both version 10 and 11 models for training, to ease the transition to the new solubility model
- The ability to use multiple accuracy extender training files enabling users to combine and extend training across projects, and extend the applicability of training in different modules
- Support for import of SMILES into PhysChem Accuracy Extender
Other improvements to ACD/Solubility DB include:
- Push data directly from Table View to Microsoft Excel at the click of a button
- Addition of 8 descriptors to the PhysChem History which depict the molecular composition of molecules
- Automatically remove explicit hydrogen atoms from structures on import of a file
- Search specific text in both user data and/or notes in database records
Download a PDF copy of the expanded details on What's New with ACD/Solubility DB, or contact your Account Manager or Distributor.
The key new features of ACD/Solubility DB presented below are:
- Expansion of the internal database
- Enhancement of the fragment-breaking algorithm
Expansion of the internal database by >20%
At ACD/Labs we are committed to keeping the internal databases of our prediction software current and up-to-date by continually adding reliable literature values. In keeping with this, the internal solubility database has been expanded by more than 20% with the addition of 1053 experimentally measured solubility values from current literature. The enhanced reference database now includes 6105 compounds.
Enhancement of the algorithm
Version 10.0 development has addressed limitations of the fragment-breaking algorithm to allow the solubility profile of larger molecules and more complex structures (such as fused ring-systems) to be calculated. Fragmentation of molecules has now been focused around reaction centers (significant in the calculation of pKa, upon which solubility values can rely heavily) to minimize the number of fragments generated for a given system.
Among the key new features of ACD/Solubility DB presented below, are:
- Improvements in solubility prediction
- Solubility training option with experimental solubility data, and
- Greatly improved database handling.
New Algorithm for Solubility Prediction
Introduction of a new fragment-dependent algorithm resulted in the increase of the overall accuracy of solubility prediction, and the addition of a software training capability with the experimental solubility data. This new algorithm represents a third generation of solubility modeling efforts for ACD/Labs. In this release, our focus was on:
- Introduction of a different model for drug-like compounds to complement the existing algorithm.
- Expansion of the training set.
- Improved predictions for zwitterions.
- Ability to train with the experimental intrinsic solubility values, similar to previous implementation for pKa and logP calculations.
New training capability allows users to enhance the solubility prediction for novel or unique chemical classes by training solubility calculation algorithms with experimental solubility values. Solubility training was previously only available through algorithm training of the underlying pKa and logP values. Starting with version 9.0, intrinsic solubility values can be used in training.
New ACD/PhysChem Accuracy Extender (Solubility Module) offers advanced capabilities for algorithm training with experimental data. Read more here.
Augmented pKa prediction algorithm resulted in the improved accuracy of solubility pH profiles and a graph outlining the distribution of ionic forms for molecules that contain several ionization centers with close pKa values, in which the dissociation path is often ambiguous.
New Database Interface and Structure
The new database enhances software training capabilities and simplifies data handling in multi-product installations. The database window now allows users to:
- View all of the properties at once.
- Include and exclude specific values into algorithm training.
- Create graphs and plots and annotate them to visualize the data values present in the database.
- Add formulas and script to the User Data fields.
- Create reports using customer's own custom templates.
- Add solubility values as intrinsic solubility or as solubility at pH. In the second case, intrinsic solubility will be calculated and marked in blue in the table.
- Add results from user's computational tools to the ACD/Labs databases.
Changes to ACD/Solubility DB, version 8.0, include significantly increased speed of calculation, added Assisted Multi-Generation capability (AMG), reworked database of solubility values, and numerous enhancements to the interface, searching, and reporting options.
Considerably improved speed of calculation:
- The calculations are now, on average, four times faster!
New AMG capability: multiple properties, multiple compounds, on the desktop:
- For small libraries and other data sets with fewer than 999 structures, you can now import and process SDfiles using the standalone desktop ACD/Solubility DB package. Calculated properties can be exported back to the SDfile, with all of the original SDfile data intact, or reported in accordance with the customizable ACD/ChemSketch templates. You can also review all of the calculation protocols one-by-one through the usual desktop ACD/Labs interface. Essential for library design.
- For customers who have several PhysChem modules, it is now possible to calculate all properties for that compound at once, view them on the screen (History Window), and export them to an SDfile.
Support of quaternary ammonium salts.
Revamped database of Solubility values now includes over 5000 chemical structures. Users of earlier versions of ACD/Solubility DB can merge the new database with their databases.
Augmented database search options:
- View and save all search operations and list manipulations performed on the currently open database. Store these search history macros as buttons on the Top toolbar for quick repetitive execution.
- Perform case-sensitive text searches through User Notes and User Fields.
- Search by Screen Form by either creating a search query on the screen, or using the current record as your search query.
Additional programming options for your database:
- With the newly implemented Add-on Organizer, use customized Add-ons to resolve specific challenges in your work, for example, remove the counter-ions from salts drawn with a covalent bond or in ionic form to unify input for the predictions.
Customized reports can either be generated from your single result (Results window), or use the database custom-designed Screen Form as a template (History window). You can also report all calculated properties at once from the new History Window. Print or save as ACD/ChemSketch or Adobe PDF files.
Unsurpassed system training abilities for pKa predictions are implemented through NEW ACD/pKa Accuracy Extender, to complement the logP expert customization option (ACD/LogP Accuracy Extender) added in version 7.0. Expert chemists can now use their own measurements to create new Hammett equations for their novel chemical classes. Resulting higher accuracy of pKa predictions has a direct impact on the calculation of pH-dependent solubility profile.
Consistent with our commitment to continually update and improve our software products according to our user's needs and requirements, we introduced a number of new features to version 7.0 of ACD/Solubility DB.
This new version allows you to:
- Create your own Solubility Databases to store your experimental solubility values.
- Export the Solubility pH profile to ACD/Curve Manager. This new feature allows users to compare predicted curves to the experimental profiles and to store predicted and experimental data in one structure-searchable database.
- Check and choose Tautomeric forms. Prior to prediction, the user can check the existence of tautomeric forms, and choose which structure to use for prediction. The option can be switched OFF and ON.
- Report your calculations:
- Send a History file or a User Database record by e-mail from the toolbar menu.
- Make reports at a click of a button. Select the data to send to the report and use ACD/ChemSketch reporting tools to create a finished document.
Numerous helpful features are added to help create and manage User Databases:
- Avoid mistakes when entering new data into a User Database. You may choose to display a warning if the difference between the entered value and the calculated value is greater than a pre-determined variation set by the user.
- Compare databases, merge them, or find common intersecting records in the User Databases. Track the success of the merging operation with the Merge Information log.
- Search by multiple User Data fields simultaneously. Boolean logic operators such as 'AND' and 'OR', or their combination, allow users to perform advanced searches.
- Highlight the searched substructure. Visually identify the fragments of the molecule that were requested in the original query. Choose the color of display and switch the option on and off.
- Search the database for ALL tautomeric forms. Software checks if the structure is available in different tautomeric forms and suggests the search options:
- Enhanced SDfile import/export. Allows user to preview the field names in an SDfile during import to allow mapping to existing fields in the database.
Improved pKa prediction algorithm for a number of chemical classes resulted in higher accuracy of Solubility pH-profile calculations.
ACD/Solubility DB v. 6.0 features improved prediction and training capabilities and offers a number of new exciting options:
- Similarity Search: in addition to the existing structure and sub-structure searches, you can now search the database to discover compounds with similar chemical structures;
- Enhanced accuracy of prediction: since pKa and logP calculations are the underlying tools for ACD/logD and solubility calculations, recent improvements in pKa and logP algorithms resulted in better accuracy of both solubility vs. pH and the intrinsic solubility predictions;
- Enhanced capabilities for training the algorithms with experimental data: pKa and logP prediction engines (ACD/pKa DB and ACD/logP DB packages) can now be simultaneously trained with multiple databases to improve the accuracy of calculations for user's datasets and to reduce training time; and
- The ACD/Solubility DB database can now be viewed in table format.
|
