|
Bio-pharmaceutical organizations face the challenge of trying to overcome the
very low probability associated with taking a clinical candidate drug successfully
to the market. Drug discovery and development has traditionally been a
long and expensive process with the costs of bringing a single drug successfully
through to the market estimated at $350-500 million. Currently, only
1 in 12 drug candidates entering clinical development make it through successfully
as a marketed drug. Non-optimal ADME (Absorption, Distribution, Metabolism, Elimination)
characteristics are responsible for greater than 30% of the failures in clinical
development resulting in increased costs and time-to-market of a drug. Having the
ability to accurately predict the ADME characteristics or the pharmacokinetics of
a drug candidate at the concept stage will allow pharmaceutical organizations
to design and develop candidates that are most likely to succeed in the clinic,
and avoid the costs associated with moving forward less promising candidates.
Tailor-made ADME models
Strand’s Predictive ADME tool can be tailored in for use
in several ways. The standard option for any user is a model that is built on
a general dataset of compounds. In such a situation, the training and test
sets used to construct the model are chosen so that they contain several types
and classes of drugs. Such models will be widely applicable for compound sets
whose pharmacokinetic parameters vary over a large range of values. The
user also has the option of selecting models that are trained with special
classes of datasets. For example, drugs specific for a particular disease or
mode of action such as antibiotics, NSAIDs, anti-neoplastic agents, etc can be selected.
We offer the user the option of tailoring the models to represent
constraints on the pharmacokinetic parameters. For example, a model tuned to
identify all compounds that have half-lives less than twelve hours can be
created for the user. For the latter two more specialized types of models,
appropriate datasets from the drug library will be used in training and validation.
The system is designed to allow medicinal chemists to understand the contribution
of a particular feature of a molecule to a particular pharmacokinetic process,
i.e. what the relationship between chemical structure and disposition in the body is,
thereby allowing them to design molecules that have favorable properties.
In addition it allows pharmaceutical organizations to prioritize their hits
and leads based on ADME and pick the ones that are the most suitable for a particular disease.
No need to reveal any confidential info
Strand has designed a method to provide predictive ADME consulting services
without the need for you to reveal the chemistry of your confidential molecules.
Strand had developed a special software program for this purpose called
Descriptor Generator.
How the Descriptor Generator works:
Strand provides you with the Descriptor Generator program.
It accepts the 3D structure information of a molecule or a set of molecules in the SD format.
It’s output is a tab-separated file that contains 1054 descriptors
calculated for each of the molecules provided.
The descriptors capture 2D and 3D properties of the molecule
including characteristics such as surface area, lipophilicity, hydrophilicity,
atom centred fragments, connectivity, etc. Various classes of specialized
descriptors including HOMA, RCI, HOMT. RDF, MOR, GETAWAY, WHIM, etc are computed.
608 descriptors capture 3D characteristics while the rest use 1D/2D or connectivity information.
The Descriptor Generator runs on both Windows and Linux platforms.
Contact us
Please contact us to know how we can help you in realizing
successful drugs, virtually now! |
