About UsOur VisionOur TeamIn the NewsEventsFAQContact UsSaty Informed



Frequently asked questions about...











What are the origins of Coalesix?

Coalesix Inc. was created as a result of a successful research project undertaken by Icosystem (www.icosystem.com) for one of the premiere pharmaceutical companies. Icosystem is a world leader in Complexity Science and Evolutionary Computing. Icosystem proposed applying their Interactive Evolutionary Computing expertise to Lead Optimization/ Candidate Design with the pharmaceutical company. Interactive Evolutionary Computing is well suited for the challenge of lead optimization/generation as it is able to take into consideration the computational aspects of drug discovery as well as the intuitive nature of medicinal chemistry candidate design.

The project turned out so well that Icosystem decided to spin-out Coalesix to concentrate on further developing the technology and providing it to other pharma and biotech companies worldwide.

Coalesix is headquartered in Cambridge, MA and was incorporated in late 2004. Its first funding round occurred in December 2004.


Who are the key management of Coalesix?

Jim Wikel, CTO
Dr. Ihsan Ecemis, VP
Chris Bingham, VP


Who are Coalesix’s financial backers?

Among the company’s investors are the original pharmaceutical collaborator, a well-known technology entrepreneur, and
Icosystem Corp.

What is the key focus of the company?

The focus of Coalesix is on increasing the efficiency of the Lead-to-Candidate phase of drug discovery. This is achieved through use of the company’s Mobius platform that identifies a number of potential series and optimizes multiple molecular and pharmacological properties in silico before expensive resources are committed to the synthesis of the compounds and subsequent biological evaluations.

Increase the number of quality shots on goal!



What does Coalesix have to offer to the
Lead Optimization community?

Mobius is a desk-top-based interactive, collaborative environment that uses input from existing predictive models, human expertise and proprietary multi-parameter search algorithms and Interactive Evolutionary Computing (IEC) to design numerous Lead series and identify alternatives for synthesis that satisfy multiple objectives.

The end result is more, better candidates faster & easier.


What is IEC?

Interactive Evolutionary Computing is a specialized optimization method from the more general class known as evolutionary algorithms (EAs) or genetic algorithms (GAs).

IEC uses human evaluation in the optimization system, both to guide the direction of the search and as part of the fitness function. This enables criteria that may not be readily expressible by an algorithm (such as synthetic tractability) to be included during the search process, ratehr than ignored or applied independently at the end of the search, dragging the final solution away from the Pareto fronteir.  


How does this apply to Drug Discovery?

The Lead-to-Candidate phase of drug discovery is a multi-criteria optimization problem in which some of the criteria cannot be easily evaluated by computational approaches. The users of Mobius select a number of computational tools that are relevant to their project and assign relative priorities to the criteria that are evaluated by these tools.

Mobius then explores a potentially vast space of possible structures, also defined by the user. During this exploration, Mobius uses the selected computational tools to evaluate potential structures, generating data about the predicted performance of each structure along the way.

Eventually Mobius presents the user with a selection of structures that appear to satisfy the priorities set by the user. The user then evaluates these structures and provides direct feedback on the attractiveness – or lack of – to Mobius. The user may decide to change the criteria priorities and the search continues until alternative structures of sufficient potential interest are identified.


What are the overall advantages MOBIUS provides
to the company?

  • Strengthens bond between CompChem & MedChem therefore fostering beneficial interaction between the two groups.
  • Helps leverage and optimize all resources: computational, medicinal, synthetic, etc.
  • Greatly expands the number of ideas that are generated and considered and increases the amount of in silico pre-screening, hence delivering more alternative candidates with potentially reduced synthetic effort i.e. “more shots on goal”.
  • Increasing the number of alternatives that are considered for the clinic represents a significant improvement in the quality of the company’s decision making - ultimately leading to a reduction in clinical attrition.



What specific benefits does the MOBIUS environment
offer the Medicinal Chemist?

  • Plays to MedChem’s strengths. Mobius allows MedChem to stay within realm of intuition/experience/insight while leveraging power of computational results/methods in order to identify multiple potential solutions in a large search space.
  • Expands discovery universe. Mobius does not exhaustively enumerate the search space in order to identify optimized solutions, so its objective-oriented search will allow scientists to evaluate a bigger, more diverse search space than more serial, iterative approaches.
  • Rapid generation/analysis of ideas. Mobius is designed to ignite the creative spark that is frequently the key to successful medicinal chemistry. It does this by generating ideas that are consistent with the general guidelines communicated by the user, but which are not constrained by inherent human bias or conservatism.
  • Allows for parallel exploration of new alternatives. As new ideas take shape, the parallelization aspect of Mobius maximizes the number of “variations on theme”/twists/alternatives that can be explored to produce the most attractive developments on each idea, leading to more, better candidates.


How easy is the environment to interact with?

The interface presented to the user is a “windows like” portal. The interface was designed to be novel in appearance while maintaining and intuitive interaction with the user. The interface is very visual in presenting information to the user.



What specific benefits does the MOBIUS environment offer the Computational Chemist?

Provides venue for CompChem to express their methods, models, ideas and to achieve a widely recognized impact on projects. Computational chemistry experts able to refine their tools hence tools are better able to meet needs of medicinal chemist.


How does CompChem work with Mobius?

The Computational Chemist works with the project team to select the individual computational tools and models that are best suited to the specific project. For each computational element, the Computational Chemist can advise on the appropriate settings for the range of possible returned values and how Mobius should interpret them.

As the Computational Chemist develops newer versions of models and tools over the course of a project, these newer versions are automatically taken up by Mobius with no interruption to the search.

As Mobius explores the relevant structure space, it is generating significant amounts of data in relation to the predicted performance of structures. By using Mobius on his own account – or by analyzing the data generated by others – the Computational chemist can also leverage his intuition and experience by identifying various structural components that appear to be driving performance and feed these insights back into the search process.


What type of computational models can be accessed by the environment?

MOBIUS allows for all the company’s predictive models to be incorporated into the environment. These range from commercially licensed software such as Accelrys, Tripos, OpenEye, etc to in-house developed code. Essentially any program that has a command line interface is a model that can be made available in the Mobius environment.



From an IT perspective, how is Mobius constructed?

Mobius has four key components: the User Interface, the Search and Design Engine, the Activity and Results Database, and the Predictive Models.  The User Interface is a Java-based client application installed on end users' PCs.  The Activity and Results Database is a SQL database such as Oracle or MySQL.  The Predictive Models are computational models used to evaluate compounds, and can be either developed by the customer, licensed from a third party, or supplied by Coalesix.  The Search and Design Engine is Java-based server that communicates with the client and models and stores the results in the database.


Is it easy to get MOBIUS up and running?

Yes.  Mobius supports a variety of flexible deployment options. In the simplest and quickest configuration, all Mobius components can be installed in seconds on a single PC as an integrated application.  In other configurations, different components can reside on dedicated servers, for higher performance or to leverage existing infrastructure, such as your Oracle database or computational models.


Do I need to perform extensive programming work in order to get my existing computational software to work with the environment? Where do these models have to be located?

No.  Mobius can integrate easily with any predictive model can that be accessed via a web service or on the command-line, in most cases with minimal or no custom programming. The models can be located on the same server as the Mobius Engine, or on a different server (or cluster of servers) accessed over the network.



How do I contact Coalesix in order to get more information on Coalesix and to set up a presentation of MOBIUS?

Telephone: 1-888-MY-MOBIUS (1-888-696-6248)
Email: correspond@coalesix.com


© 2004-2007 Coalesix Inc. All rights reserved.