Time Series Prediction Approaches

Time Series Journal

Subscribe to Time Series Journal: eMailAlertsEmail Alerts newslettersWeekly Newsletters
Get Time Series Journal: homepageHomepage mobileMobile rssRSS facebookFacebook twitterTwitter linkedinLinkedIn

Time Series Journal Authors: Jason Bloomberg, Progress Blog, SmartBear Blog, APM Blog, Jnan Dash

Related Topics: Time Series Journal, Pharmaceutical News

Time Series Journal: Article

Talking with Sharon Nunes

Talking with Sharon Nunes

Jack Martin, editor-in-chief of WSDJ, recently spoke with Dr. Sharon Nunes, IBM's director of Life Sciences Solution development, about how pharmaceutical companies are handling the rapid growth of information in drug development.

WSDJ: How did IBM get involved in the pharmaceutical industries and what type of problem were you trying to solve?
Sharon Nunes: In the mid '90s, IBM Research was collaborating on a few joint projects with pharmaceutical companies in the area of drug design and in the area of creating algorithms for pattern discovery and pattern matching. These algorithms helped researchers find patterns in the gene sequences that were coming out of the Human Genome project. There wasn't a huge amount of data at that time compared to what we have today, but we realized that they were generating huge amounts of data. IBM has always been engaged with pharmaceutical companies on the business side, but this was a new, exciting opportunity for us - IBM Research could create new tools to help analyze data being generated around the Human Genome project.

In the late '90s IBM research formally launched its Computational Biology Center, which now has about 50 people around the world doing various aspects of research related to the computational challenges of pharmaceutical companies. Today we also have a large business organization with almost 200 people worldwide in the core group, as well as another 500 in the IBM brand groups, who are dedicated to life sciences activity. I manage the worldwide solutions development activity, which has an extended team of about 100 people. Our focus in Life Sciences is to deliver solutions to the pharmaceutical and biotech industries. Our solutions target high-performance infrastructure, data and knowledge management, and clinical development/regulatory compliance.

WSDJ: What are some of the key issues that you're trying to solve for pharmaceuticals, things they are facing today?
SN: Pharmaceutical companies are faced with an explosion of information, more so within the last five to seven years. Let me give you some specific data points. There is a technology called microarray analysis; you can get 40,000 data points off one of these microarrays. Think of a silicon chip because the technology is similar, but this has DNA strands on it; you can look at the interaction of DNA strands with a possible drug for example.

There are also many public databases that hold gene sequences and protein sequences and protein structures. One of the more popular ones, GenBeg, is growing at the rate of almost a million entries an hour. In addition, pharmaceutical companies and biotech companies themselves are generating millions of data points in lab experiments every week, whether it's mass spec experiments that they are running in the lab to do some protein analysis or they are doing some gene sequences themselves. You need sophisticated computer programs to look at the structures of these proteins and the possible interactions of the proteins with other proteins in the body or with proteins and drugs. You need huge amounts of computer storage to hold all that information, and you need smart analytical programs to make sense out of that data and to allow you to access the right data for your business decision. So, there's a real explosion of information around the Genome Project and the work that's come along subsequently relating to proteins and protein structures.

Proteins are really important because it's the interaction of drugs with the proteins in the body that helps to either minimize the effects of diseases or to cure them. Typically, when the protein malfunctions, you see the onset of disease. It's important for researchers to understand the effects of proteins in the body, and the pharmaceutical company will develop drugs targeted at a specific protein to react in a certain way to either cure a disease or to minimize its effects. The explosion of information is one of the challenges as companies grow, and there has been a tremendous number of mergers and acquisitions in the last several years with the pharmaceutical companies.

Management of information, sharing of information, collaboration, and knowledge management are all key issues that are challenging our customers. There are many "point applications," and we are hearing that integration of data and integration of applications is another challenge for our customers. The IBM Life Sciences Framework is IBM's solution to this problem that is so critical to our customers.

WSDJ: What are the two biggest things keeping pharmaceutical companies from being successful today?
SN: One of them, I really think, goes back to what I said regarding the sharing of information and collaboration. One of our customers said, "I have nine labs around the world. I know I have researchers in all of those labs who are probably working on similar projects. How do I get those researchers to find each other and to communicate with each other and to share information?" We believe that with IBM's Life Sciences Framework, the WebSphere Portal, for example, can make it easier for them to share information. We have what we think are cool tools to enable finding the right resources and organizations automatically that really can help the customers solve some of those problems.

Another key challenge relates to data integration and application integration because each customer we deal with has a number of favorite applications. There are thousands of applications in the biology and chemistry domains that our pharmaceutical customers like to use. One of the challenges is how to get the chemistry data to talk to the biology data as you proceed down the drug development process. How do you reach back and get information from the beginning of the pipeline that may be critical for you to make a good decision in toxicology, for example, testing the toxicity of a drug in either animal or human trials.

One of the things we have done in our framework is focus on open standards. This is very much in line with the IBM e-business applications framework. We call it the Life Sciences Framework and it focuses on open standards, and an open, modular architecture. WebSphere and WebSphere Portal are both key components of it. We're looking at integration of these multiple applications through XML, which is becoming very much a standard in the life sciences space. We think that these technologies are key for enabling our customers to follow the flow of information through their enterprise pipeline. As you move from biology to chemistry to clinical domains during drug development, you want to be sure you can readily retrieve data from any part of those processes. We believe it is critical to have open standards and Web services, which will allow any application to talk to another application.

WSDJ: Does this accelerate their time to discovery?
SN: We believe it can make a tremendous difference. We've used the e-business applications framework in many other domains and have information from other areas where it has enhanced productivity. We have developed some prototypes and we are actually in pilots with three customers right now using various aspects of the IBM Life Sciences Framework. We've simplified a lot of the operations into a kind of automated back-end script that enables you to automate a lot of the manual processes. Now researchers might say, "You know I can't have everything automated because there are certain points where I need to add some of my own domain expertise." IBM has a huge number of domain experts and chemists/biologists with industry experience. We have consulted with them during the development of the Life Sciences Framework and, with their guidance, have developed a prototype that automates as much as possible, but also presents options for decision-making to the scientists at certain points along the development process. We have a combination of human intervention with background automation, which simplifies a lot of the manual processes that have been used in the past.

WSDJ: Do you have any sense of how much time it saves them from a percentage standpoint?
SN: We have just implemented some early prototypes. It's hard to make a concrete prediction, but one of our customers has implemented a piece of this framework in their business environment. We've heard from them that a process which formerly took them about a week now takes them just a couple of hours. To me, that's a significant productivity increase if you can cut off four or five days from part of your development cycle. That's a tremendous savings in time of people and time to market as you start to march down the development path.

WSDJ: Why is this approach better than the competition's and how is it unique?
SN: We believe that IBM has a true end-to-end, enterprise-based framework. We also think that we bring in the best applications through an open framework so our customers can decide what applications they want to use as their front end to integrate on top of our framework. We do not require that all of the applications run just on an IBM platform because if you can integrate them with XML or Web services, they can run on competing platforms. We don't like to encourage that, but of course, that's the reality of most of this industry.

We think that we offer more flexibility to our customers by offering this open environment and we also believe that through our data integration software called DiscoveryLink - which is part of our framework - that we enable our customers to integrate heterogeneous data as well. The data does not all have to be in a relational format. It can be in a spreadsheet on someone's desktop, it can be in a flat file, it can be data that's accessed over the Web, or it can be in a relational format. It can even be in a competing relational database format, not just in our DB2 format. We believe that those two factors - access to heterogeneous information and the flexibility of integrating any application into our framework - really differentiate us from any of the competing systems that exist today.

WSDJ: Are you the only one on the playing field that's based on open systems right now? Is that what makes this so special compared to anything that a pharmaceutical company can do elsewhere?
SN: Well, the IBM Life Sciences Framework is open and modular. We believe there's a tremendous amount of flexibility. We don't run on a closed proprietary system; we think that we offer the best openness and flexibility to our customers compared to any other systems out there today. In addition, our IBM Life Sciences Framework really is an embodiment of the I3C standards, which are focused on the interoperability of systems and infrastructures based on the needs of the life sciences and pharmaceutical industries. We believe that because of this openness, flexibility, and focus on standards, IBM has an incredibly powerful solution for our customers.

I think one of the reasons why something like this framework and enterprise-based system is important is when we talk to our customers about their own costs, it takes 10 to 12 years to develop a drug, including the clinical trials and approval through the FDA. Every day that they can get a drug to market sooner means a million dollars of opportunity in their pockets. There is a tremendous financial incentive to accelerate the drug discovery process as much as possible.

WSDJ: What would a pharmaceutical company find over the whole 12-year period it takes to go from concept to drug approval by the Food and Drug Administration?
SN: The numbers that came out of a study about six months ago indicate that the cost for developing a drug is somewhere around $800 million. The actual cost is debatable, but I think that most people agree that it's somewhere between $500 million and $800 million to develop a drug. The process for developing a drug entails three or four key areas of development. One is at the beginning when you are identifying the biological target in the body. There is a lot of biology- and genetics-related research there.

The second phase is where you do the drug development itself, taking a potential drug molecule and understanding whether it's an appropriate drug and whether it actually affects or interacts with the particular target in the body that you have identified from your first phase of study.

Then you have to develop the drug and make sure that it's absorbed into the body appropriately, that it metabolizes appropriately, that there are no (or minimal) toxicity effects. Then you go into a series of clinical trials, starting with trials on animals, then proceeding into trials on human subjects.

The whole process is somewhere around 12 years on average. When you take into account the expense and the number of years of development, there's a tremendous incentive to decrease the cost, even if by a factor of 10%. Ten percent is a tremendous savings for these pharmaceutical companies when you take into account the overall time and cost.

One of the advantages to the framework is that we've developed in a very modular fashion. If you can develop this on a departmental level and get immediate results with it, you can scale that to a larger group level and you can look at it from one location and from an enterprise worldwide implementation. The investment depends on the amount of time and the amount of resources that are needed to help solve some of these problems.

More Stories By Jacques Martin

Jack Martin, editor-in-chief of WebSphere Journal, is cofounder and CEO of Simplex Knowledge Company (publisher of Sarbanes-Oxley Compliance Journal http://www.s-ox.com), an Internet software boutique specializing in WebSphere development. Simplex developed the first remote video transmission system designed specifically for childcare centers, which received worldwide media attention, and the world's first diagnostic quality ultrasound broadcast system. Jack is co-author of Understanding WebSphere, from Prentice Hall.

Comments (0)

Share your thoughts on this story.

Add your comment
You must be signed in to add a comment. Sign-in | Register

In accordance with our Comment Policy, we encourage comments that are on topic, relevant and to-the-point. We will remove comments that include profanity, personal attacks, racial slurs, threats of violence, or other inappropriate material that violates our Terms and Conditions, and will block users who make repeated violations. We ask all readers to expect diversity of opinion and to treat one another with dignity and respect.