Making informed decisions

Using informatics solutions to manage data in the drug discovery process. With STARLIMS Corporation and TeraDisc

“The software has sown promising results in terms of achieving very accurate binding predictions”
-Ed Addison, TeraDisc

NGP. What technologies can companies use to help reduce the cost of drug discovery?
Simon Wood. It is clear that many different informatics solutions are used to manage the huge amounts of data generated in the drug discovery process. This data and information are key to making informed management decisions for individual discovery projects, especially as far as go/no go or fail early decisions are concerned, and these are key to efficient use of resources, including people and money. However, this information is often fragmented and isolated throughout the discovery organisation. Laboratory informatics solutions, such as scientific data management systems (SDMS), can unify this data into a single source, and make it available to the people and systems that need it. This supports efficient decision-making, therefore helping control costs.

Ed Addison. At TeraDisc, we are reducing the cost of drug discovery using a QM/MM algorithms with intelligent search over molecular space using a 2,200 CPU high performance computing cluster. The software is patent pending and university generated and has shown promising results in terms of achieving very accurate binding predictions for both small molecules and peptides. We are able to discover lead compounds by searching theoretical space rather than an existing molecular library. The results are a highly focused library of good binding leads generated ‘in silico’ in just a few months using a tool known as ‘Inverse Design’. We do this for clients as a service.

NGP. With looming patent expiries, pharmaceutical companies need to be more efficient in R&D to stay ahead of the competition. What tools can they use to help achieve streamline their drug discovery process?
EA. The pharmaceutical companies that are able to use high performance computing as a preliminary step in discovery – for virtual screening, target validation, toxicity assessment, or mechanism of action studies – will be better positioned to save time and money in the discovery process.

Examples of tools running on high performance computers that help with drug discovery costs are biomarker extraction algorithms, target simulation, virtual screening, and toxicity evaluations.

SW. From an informatics point of view, the tools that help companies cut the cost of drug discovery are the same as the tools that will help streamline the drug discovery process by facilitating the decision making process. So again it comes down to those systems that will provide unified access to the required data no matter where it comes from or what systems generate it. An SDMS, for example, should be capable of taking project data from multiple sources, extracting key data, storing it in a format-neutral way (which also preserves long-term value of the data) and making it available for searching and reporting

NGP. Drug firms are facing increasing demands for compliance from regulatory bodies. What tools can they use to help them meet these requirements?
SW. Perhaps the most important tools that will help drug firms meet their regulatory requirements are the ones that will help them implement and manage the traceability requirements of those regulations. Some estimates put the amount of time spent on regulatory and compliance activities within the lab to be as high as 70 percent. Clearly, any informatics systems implemented in the lab must support the regulatory needs to try and reduce this burden. Perhaps the most obvious example of this is the support of electronic signatures in line with FDA 21CFR Part 11 regulations. However, systems such as LIMS will also support other regulatory requirements – including managing instrument maintenance and analyst training and certification, inventory management and chain of custody records. They will also provide an automatic audit trail of actions taken and changes made. Using informatics solutions to manage regulatory needs not only eliminates possible missing records, where lab staff forget to record the required information, but also makes the retrieval of regulatory data easier.   

EA. One tool that we are developing is a high performance PK/PD algorithm that runs on high performance computers and that is capable of analysing clinical trials outcomes on very large trials with many parameters, even high content data. Our tool is based on a Bayesian approach. This will enable drug developers to provide statistical proof of fine grain outcomes useful for personalised medicine, biomarker identification, and/or risk assessment.

NGP. How do you see the future of drug discovery developing in the next few years?
EA. We believe the market must prepare for a transition to a greater use of life science computing to be more efficient, and from a business point of view the market must address the transition to personalised medicine. This means that more drugs with smaller market sizes must be developed quicker.

Despite the uncertainties surrounding precise cost and time estimates in the field of personalised medicine, various impacts on the drug development process, pipeline and industry value chain have been identified as more powerful and safer drugs, due to genetic specificity and ability to choose dosing based on biomarkers. This also requires the provision of genetic and other diagnostics to test biomarkers, which has led to an expected improvement in drug discovery and development through increasing knowledge of genomics and bioinformatics.

Another impact has been the increased need for IT in the drug development, discovery, approval and clinical administration process, along with a decrease in pharmaceutical pricing and health care costs, due to lower development costs

The onslaught of personalised medicine in the marketplace represents a classic discontinuity.  It is currently in the ‘incubation’ period. However, when personalised medicine reaches the rapid growth period, there will be a paradigm shifting affect in the industry that accounts for all the factors above. While the specific timing and magnitudes of the metrics involved are not known yet, the gradual industry shake up expected to occur will have a significant impact on the business models, pipelines, and growth rates of major pharmaceutical companies. Embracing life sciences computing models will greatly improve the efficiency of this transition.

SW. The pressure on drug companies to streamline their pipeline and bring candidate compounds to market quicker will only continue; and the pressure will be on drug discovery organisations to identify high potential compounds to deliver to the rest of the value chain. Continued consolidation within the industry in terms of mergers and acquisitions and the continued development of collaborative projects will all affect drug discovery organisations. The reliance on IT solutions will increase, but organisations will be looking for unified systems and solutions that allow them to concentrate on realising the true value of the data and information that exists within the whole organisation, as opposed to just managing the data produced. The ability to deliver the data and information produced throughout the organisation to the people and systems that require it will be of paramount importance.
 
Dr. Simon Wood is Executive Director, Marketing and Education at STARLIMS Corporation. A leading authority in laboratory informatics, Simon is a popular lecturer on system implementation and laboratory IT. His prior experience includes establishing, for Thermo LabSystems, the industry’s largest LIMS implementation team. Simon holds a PhD from Sheffield University.

Ed Addison is an established serial entrepreneur and CEO of TeraDisc, a company that provides high performance computing applications for drug discovery.  Named “Entrepreneur of the Year” after achieving #51 on the National Fast 500 in 1994, Ed is also an Adjunct Professor of Bioinformatics at Johns Hopkins University.