| MDL QSAR is a comprehensive QSAR modeling
system that enables scientists to establish reliable quantitative
structure-activity and structure-property relationships, create new
calculators for in silico screening, and generate new compound
libraries based on results—accelerating the discovery of lead compounds
in drug and agrochemical research. By providing powerful computational
tools in a user-friendly environment, MDL QSAR supports a highly effective
drug discovery workflow across all project team disciplines.
Highlights:
One-stop Analysis
Designed as a turnkey solution, MDL QSAR provides all the tools necessary
to perform the most demanding QSAR analyses in an easily accessible,
Windows-based workspace.
Built-in and Custom-tailored Descriptor Sets
MDL QSAR performs rapid processing of thousands of molecules using over
400 built-in 2D and 3D molecular descriptor calculators, including
molecular property descriptors, total topological descriptors, and E-state
indices enabling researchers to model many activities and properties.
Additionally, MDL QSAR makes it easy to import your own or third-party
descriptors or use measured endpoints as custom descriptors.
Versatile Statistical Routines with Feedback
Qualifiers
MDL QSAR employs a powerful Genetic Algorithm that automatically chooses
the best set of descriptors to create a model from a given dataset. The
model-building workspace provides numerous built-in statistical tools
including multi-linear regression, principal component analysis, partial
least squares, all possible subset analysis, stepwise linear regression,
and cluster-analysis.
Reliable, Easy-to-Use QSAR Modeling
MDL QSAR includes a unique "One-Touch" model-building tool that
lets researchers construct reliable QSAR models in seconds—yielding
fast, reliable results when exploring large data sets for correlations or
testing hypotheses. Researchers can then use more advanced features to
further refine and analyze their models. MDL QSAR generates precise,
easy-to-interpret, model-quality metrics and explains various linear and
cross validation metrics to ensure the correct interpretation of models.
The application then presents molecular structures and QSAR results using
state-of-the-art, interactive graphic displays. This ease of use means
that an entire project team can be creatively involved in the discovery
process.
Flexible Screening
MDL QSAR makes it easy for researchers to qualify new compounds by
screening thousands of unknowns in batch mode against a given model, while
automatically calculating the descriptor-space distance for each new
compound from that of the model set to gauge the validity of the results.
Build and Screen New Libraries Based on Results
MDL QSAR lets researchers designate R-Groups on a scaffold, then use a
Substituent Library to generate and automatically screen new combinatorial
libraries. Flexible design also lets you add new structures to the
Substituent Library for project-specific analyses.
Rapid Database Search and Retrieval
MDL QSAR offers relational database and storage features that are ideal
for building and managing molecule libraries for a QSAR analysis. Through
the use of descriptor, substructure, or structural similarity filters, MDL
QSAR makes it easy to identify the best compounds to build and refine your
model.
Last Updated 28-Jun-2002 |