|
2008 upgrades are coming soon
PrologP
Our new PrologP prediction model will include an extended fragment collection and a more ellaborated neural network algorithm. The new prediction engine considers the result of several linear and neural network-based models in order to provide a more accurate and more robust estimation of the water-octanol distribution coefficient. The method involves forming the current member of a new class of pseudo-linear algorithms, where the precision of the non-linear approaches is combined with the transparency of earlier linear methods.
pKalc
In order to enhance the accuracy of pKa predictions, several acidic and basic models have been recalculated, using more than 10,000 experimental pKa values collected from the chemical literature. Thanks to the newly inserted parameters, the upgraded version of pKalc provides significantly more accurate predictions.
MetabolExpert
The knowledge base of the MetabolExpert module of Pallas has been extended with new metabolic reactions. The extension includes a set of common metabolic reactions collected from recent scientific literature focusing on the metabolisms of toxic and drug-like organic compounds. The new version of MetabolExpert will manage the common metabolic reactions and a series of special metabolisms also.
HazardExpert
The HazardExpert knowledge base has been extended with a collection of carcinogen and mutagen groups, based on experimental in vivo data. The inserted new fragments will allow HazardExpert to recognize potencially carcinogen compound structures more effectively.
|