The Challenge

Facing payer pressure and the need to fill dry R&D pipelines, the pharmaceutical industry seeks new ways to redefine and increase its competitive edge.

Pharma continues to restructure its internal R&D departments as part of widespread cost-cutting measures to help reduce the impact of patent expiration and escalating operating costs. With fierce competition for late-stage products, in-licensing these opportunities are less accessible and driving prices upwards adversely impacting ROI. As a consequence, would-be licensees must look at either less commercially attractive late-stage drugs or earlier-stage licensing opportunities that hold true novelty and promise. Even with complex early-stage molecules, companies have to be far more creative and flexible in their approach to securing the best deal terms, with traditional licensing deals often being replaced by option arrangements or heavily back-ended deal structures - all to accommodate risk.

My recent blog posts have talked about using a semantics intelligence platform to help organizations overwhelmed by structured and unstructured data identify the telltale signs of novelty and value particularly with early-stage assets. What’s attractive is the potential to increase a company’s ability to win ahead of the competition. Large volumes of new content are generated daily, limiting progress in both search and general business information management. The reason being that text and structured data content stores (silos) are very fragmented, disconnected and difficult to access. Not to mention the information hiding in unstructured data. The result is often protracted decision-making, higher risk decisions and missed opportunities.

To succeed, pharma must rapidly piece together critical insights from multiple data silos, incompatible systems and isolated processes. Often no one tool can provide the answer. About 30 to 40 percent of a worker’s time is, in reality, actually spent searching for information that already exists within an enterprise, though they find it there less than half of the time. Are companies really efficient at finding what they want or are they fooling themselves thinking they are doing it right and that someone else will pay for their time?

The Inefficiency of Finding Assets

Quite often some of the best insights and decisions, “tacit interactions” are captured and remain in silos for emails, meeting notes, blogs, and wikis. This fuels the fragmentation of the R&D, licensing and competitive intelligence “knowledge-base”. Companies are unable to unify data let alone look across information multi-dimensionally and finding the “interesting stuff” in the meta-data to derive new insights is near impossible. Enter the issue of hidden people time. Are executives are getting paid for the information inefficiency? There are two culprits. Creating information and finding it.

First, people spend as much as 90 percent of their time creating information that already exists. In the pharma industry, whether in R&D or scouting for a potential licensing opportunity, this means running assays or studies that aren’t really needed or asset search dead ends. The result, increased cost and time…or passing up critical insights.

Secondly, as much as 50 to 75 percent of information is found by interacting with other people, not from systems that may hold the content. Here is where the trouble starts and what the industry needs to ask itself: Are we missing something important that can change drug development or in-licensing directions? What is it that “we don’t know we knew,” especially when the information and knowledgeable people resources actually exist, but cannot be discovered easily?

Why Semantics Intelligence for Asset Discovery?

That is why a process based on semantics intelligence holds promise. Unless you have a small army of research investigators, semantic intelligence may prove to be the answer. Here’s how. It’s all about disambiguation. The improved precision and recall from semantics combined with predictive analytics driven from unstructured content, linked to structured content not only can help find compounds in pre-clinical and clinical phases; it can also identify critical elements of a pharmaceutical in its chemical state and help value them on basis of their novelty and attractiveness. This expands the potential for pharmaceutical manufacturers because it provides them with two options: first, the ability to rapidly seek a known compound or entity in some stage of development and second, the capability to find an ideal chemical with a molecular structure having attributes that could be used to enhance the performance of a new drug or improve an existing one.

Value in Competitive Intelligence

If a drug company executive finds out about a new compound or formula, it can now tell if the discovery holds promise early and consistently with the platform because it addresses information search and orienteering as well as asset qualification.

The first step known as scouting and filtering is critical so the solution gets the right information in the right context to the user with a high degree of clarity. The solution can increase precision and recall as it relates to assessing new trends and opportunities, and can compare the impact to a competitor’s from a single query. Additional benefits include:

Resolving semantic mismatches across disparate data sources

  • Ontology designer
  • Model both Data and processes
  • Open RDF data store, Oracle,

Integrating a variety of data types

  • Structured, semi-structured, and unstructured data
  • Scalable to very large data sets

Finding patterns and cleanly present insights within large data volumes

  • Provide an inference and a reasoning linking the resulting data from the initial query
  • Scalable Reasoning Systems

Deploying high-value application with “discovered knowledge

  • Real-time complex events processing
  • Business application modules
  • Custom applications

Opprotunity Assessment in a Virtual Decision Environment

Once the asset is located ahead of your competitors then what? On approach is where a solution quantitatively relates real-time variations in company functional-area input to a decision to abandon or pursue a given opportunity. Picture a virtual environment that captures and applies a deep set of qualifying rules for a decision. Just as in the physical world, virtual clinical, managed markets, marketing, IP, financial, and regulatory departmental approval criteria are applied so a company knows whether to pursue or abandon an opportunity. These rules or criteria could then fed into a scoring model that can be applied to screen new opportunities based on a host of attractiveness measures. Behind the scenes, it weighs department criteria based on importance to yield a range of values that serve as the framework for a financial worthiness simulation.

Paragon brings pertinent information into the dashboard
  • Each functional area influences the financial performance of the other
  • As an assumption is changed in one area, the financial effect on the others on revenue and profit is simulated over time
  • Interactive decision-making modeling incorporates multiple interdependent variables that drive financial performance, risk and attractiveness
  • Accommodate impact of change by functional attribute to assess the financial impact on product, therapeutic area, franchise, division or entire enterprise
  • Workflow engine models change to guide selection of best performing alternative for licensing.

longer for a return, drugs licensed at the pre-clinical phase would be expected to create the maximum amount of value for the company. For instance, if there are only a few drugs (or none) to treat a certain condition asset may be more valuable because it could offer something distinctive in a market that’s not crowded with other options. On the other hand, a company may not want to enter a saturated market with a drug development because they may be able to pioneer treatments in another therapeutic area that boasts less advancement.

Semantic Technologies represent data in context enabling useful and intelligent operations to be performed on and with that data, with and potentially without human intervention. Using a semantics intelligence platform can help a pharmaceutical company pinpoint the asset and more quickly get started on the research phase and boost the development phase. It would also accelerate the licensing process as well because the system can scout what’s going on in the industry and determine where opportunities lie. While it would not accelerate the drug development lifecycle as a whole, it certainly would put it on the fast-track at the same time as better identifying risks, costs, disadvantages, and benefits.

This kind of time can be essential, bringing breakthrough medications to the market quicker than ever before. What additional values exist for a system that can foretell the viability of a major business investment?