Quantcast

Industry news that matters to you.  Learn more

2011 ISB Systems Biology and P4 Medicine Symposium, Day 1

Reading time: 10 – 16 minutes

The first Institute for Systems Biology (ISB) symposium was held in 2002. Now in its 10th year, the Systems Biology and P4 Medicine conference provides a setting for some of the world’s most influential researchers who are leading the way in applying systems biology and systems theory to medicine and health care delivery.

The symposium kicked off with welcoming remarks from ISB’s co-founder and president, Lee Hood, MD, PhD. Here are some of his comments and ideas captured from his presentation:

  • In 10 years, P4 medicine will generate billions of data points around each individual to sculpt health and disease.
  • Systems, networks cause disease, not individual genes.
  • Evolution accounts for the magnitude of biological complexity:
    • global data sets, look at all data types
    • integrate data sets
    • capture, understand dynamics (both temporal and spatial)
    • deal with the enormous signal to noise
  • Handle biological noise by using lots of samples to subtract background.
  • How do we convert data into knowledge? Models arise from a systems approach to disease. Create model from extant data, formulate hypothesis to test through experimental perturbations of the system, and then iterate the process.
  • Biological information is modular, provides key opportunities for deciphering complexity.
  • Two types of biological information: digital genome and environmental signals.
  • Blood provides a window to measuring disease. Need longitudinal data and multiple measurement, protein, RNA, DNA.
  • Key features of biomarkers in the future: multiparametric, work just for the individual (become their own control).
  • No disease is an single organ island.
  • Biological networks and molecular machines are two information structures that connect genotype/environment with phenotype.
  • Time to start thinking about the type of IT required to capture, manipulate billions of data points.

Hood then went on to describe the P4 Medicine Institute, an innovation consortium that plans to lead the emergence and adoption of healthcare of the future which will be predictive, preventive, personalized and participatory. He concluded his talk by listing four societal implications of P4 medicine: one, that it will force a revision of business plans; two, that it will digitalize medicine; three, that it will turn sharply around the escalating costs of healthcare; and four, that it will create significant wealth for those who adopt it.

George Poste, PhD, Chief Scientist at the Complex Adaptive Systems Initiative and Regents’ Professor and Del E. Webb Chair in Health Innovation at Arizona State University, then gave the opening keynote address, speaking about opportunities and challenges in the evolution of personalized medicine. Poste’s talk was quite complimentary to Hood’s, focusing on research and healthcare costs in connection to P4 medicine (a copy of his presentation can be found here). Here are some of his comments from his presentation:

  • How do we sustain innovation in era of economic constraint?
  • 20% of population generates 80% of the costs of healthcare.
  • How do we place health constraints on a generation that hasn’t ever known them?
  • Adds two more P’s to the list: Price (research, innovation, products, delivery of care) and Policy.
  • How do we move from pharmaceuticals to pharmasuitables – right RX for the right disease subtype?
  • We are in the opening era of linking disease molecular pathology to rationalize Rx.
  • Molecular profiling and segmentation of patient populations can be used to design new clinical trials.
  • It may be cheaper to do whole genome sequencing vs developing multiple Dx tests.
  • Only 121 drug labels currently contain pharmacogenomic information.
  • The relationship between genotype and phenotype is nonlinear.
  • 80k phrama drugs x 300k OTC drugs x 1000s nutraceuticals = 30B combinations. Then, add in the microbiome.
  • The evolution of drug discovery:
    • empirical screening
    • cellular/molecular pharmacology, reliance on whole cell assays retains “systems-complexity”
    • genomics and high-throughput screening of molecular targets; reductionism and elimination of systems complexity
    • mapping molecular pathways/networks, return to systems pharmacology
  • How do we map chemotypes to pathways and subnetworks for chronic, progressive disease?
  • U.S. cancer prevalence estimates in 10 years to increase 20-30%
  • Over 120,000 claimed biomarkers or biosignatures but less than 100 molecular diagnostics in clinical use or in advanced trials.
  • Predicts cancer pricing bubble is going to burst. We need to spend more money on diagnostics, less on developing therapies.
  • We spend 93% on drugs and only 7% on diagnostic research spending. We need to fix that if we a going to make progress improving health care.
  • We need to move from cost-based reimbursement to value-based reimbursement. If it’s not billable, it won’t happen.
  • Qualcomm, LG, Google aren’t coming to mobile health out of altruism. Big biz opportunity to capture, share, patient data.
  • All of us are going to have to take personal responsibility for wellness when risks are known.
  • Managing mega-data is a very complex issue: volume, scale.
  • Formidable challenge: interoperability and scalability problems.
  • Medical paternalism will be dented over time. Patients will be empowered, things like PatientsLikeMe a welcome change.
  • Social networks will eat away at medical paternalism.
  • There will be a recalibration of the physician in healthcare delivery.
  • Physics has already gone collaborative, open source. Biology has to go there too.
  • To building large-scale datasets of human disease, we must use “open-commons” community-based inputs.
  • Cultural issues, silos subvert solutions, protection of status quo. We need to question reward structures.
  • We’re going to have to get a lot more mature and address medical dollars spent at the end of life. The current practices are simply not sustainable.

Session 1: Personalized Medicine

The first session focused on personalized medicine and was chaired by Sanders Williams, President of the Gladstone Institutes, an independent, non-profit biomedical research foundation affiliated with UCSF.

First up was Anne Wojcicki from 23andMe. She spoke about consumers, genomes and research, and the power of numbers for the future of personalized medicine.

  • How many people here in this scientific audience been genotyped? (Not many hands up)
  • Healthcare lacks consumer opinion.
  • The only way to change healthcare is through the consumer’s voice.
  • Science is too slow. I decided to scrap the current system and come up with something better to empower consumers.
  • I’m from Silicon Valley. I change my website daily. I don’t want to wait!
  • 23andMe now has over 75k people in its database.
  • The initial $1000 price point to sequence DNA was too high. Consumers are not used to paying for healthcare.
  • Whole point of personalized medicine: think of things on a molecular level.
  • The 23andMe website has 180 reports on health and traits, ~44 are highly replicated.
  • You might want information in your medical record that isn’t deemed cost effective by your insurance company.
  • What do you do with genetic information? Partner with physician community to develop recommendations.
  • Drugs may have genetic labels, but docs don’t use the labels. Patients need to be informed and aware of their genetics.
  • Cites ASHG Kaufman 2010 paper, which found very little harm from a patient learning his/her genotype.
  • 58% of users learned how to improve their health, 34% indicated they are more careful with their diet, 14% exercise more and 15% changed medications.
  • The problem in the system is that companies don’t make money from people who don’t get sick.
  • Get away from the GATTACA movement. Focus on the fun of genome science.
  • A second paper from 23andMe coming out soon. We looked at 50 medical phenotypes from 20k genotyped individuals – 75% of associations could be replicated.
  • We can do research more quickly because we can get lots of people doing surveys as partners.
  • The 23andMe Parkinson’s Project with The Michael J Fox Foundation had preliminary results from 2k people in 9 months yielding a number of novel genes including a potential modifier.
  • 23andMe will start doing preprints to get data out faster.

Nicole Urban, PhD, from the Fred Hutchinson Cancer Research Center spoke next. She described her work on the identification, validation and application of biomarkers for epithelial ovarian cancer.

  • Early detection of an ovarian cancer serum marker pipeline was proposed in 1998. We spent a decade developing and validating markers.
  • In early 2008, we took part in a collaborative study evaluating prostate, lung and ovary cancer samples. Only a few markers from clinical samples performed well in preclinical samples.
  • Novel ovarian cancer markers don’t contribute to a panel. The characteristics of individual women may affect marker levels.
  • Proteomics technology has not helped to find novel markers for ovarian cancer.

Clay Marsh, MD from Ohio State University spoke about realizing the potential of P4 medicine.

  • Healthcare spending is likely to reach $4 trillion by 2016.
  • Your environment regulates DNA activity. DNA is not your destiny.
  • Behavior, environment and genetics are important for defining phenotype.
  • Future medicine goes from hospitals to individuals, participating in networks.
  • There is a connection between good health and high connectivity, lots of friends and social networks.
  • There are two main pilot projects at Ohio State: wellness and care coordination in chronic disease.
  • The average person eats 45.3 lbs of processed sugar every year. May be responsible for obesity problems in the US?
  • Doing very simple things reproducibly is key to hospitals achieving better care.
  • Participation is key in P4 medicine. Social networking is key. People spending 700 billion min/month on Facebook.
  • Consider the iPad: the hardware is just a conduit to get into the ecosystem – iTunes store with apps, entertainment, etc.
  • In complex systems, you can’t predict hubs, you have to perturb system and try to detect them.

A session panel discussion was then held where the audience could ask the speakers questions. Here are a few remarks:

  • Wojcicki: There’s a lot of misinformation out there. Some people are convinced 23andMe must be selling individual genome data. That’s not true.
  • Wojcicki: We have consumers motivated to participate. We want to help drug discovery at pharma, and research like at ISB.
  • Marsh: The IOM says it takes 17years to go from discovery to application. We need to shorten the timeline.
  • Urban: Proteomics has limitations: a small tumor is not going to create high levels of protein in the blood. Instead, maybe we should look at autoimmune proteins?

Session 2: Personalized Genome

George Church, PhD, from Harvard Medical School chaired the second session on the personalized genome. First, there was a panel discussion on current sequencing systems with Mark Stevenson (Life Technologies), Cliff Reid, PhD, (Complete Genomics) and David Bentley, PhD, (Illumina).

When asked “Where is the future of DNA sequencing?” not surprisingly there were different answers. Life Technologies sees the future in ion sequencing and semiconductor based detection systems. Complete Genomics, which only sequences whole human genomes, sees the future in cloud-based genome sequencing. Illumina sees the future in SBS chemistry. Below are some comments captured during the discussion:

  • Stevenson: Next-Gen sequencing is getting used for discovery and Sanger for the assays.
  • Reid: Two formative technologies of our era. FedEx and Internet. One moves atoms, the other moves bits.
  • Reid: Myriad and Genomic Health sell diagnostic tests for $3,000 and turnaround the data in two weeks. They set the bar.
  • Reid: we believe genomics is a big computing problem. I have more software engineers than biologists in the company.
  • Bentley: There’s been a shift from any one company taking responsibility for genome assembly and alignment to a shared public task. What we need and lack is an evolving standard for the interpretation.

The day finished up with Eric Schadt, PhD, (Pacific Biosciences) and Barrett Bready, MD, (NABsys) discussing future sequencing systems. Both companies are working on sequencing single molecules of DNA but in very different ways. Pacific Biosciences is using a semiconductor containing nanopores, through which a DNA molecule can be threaded. This creates a nanophotonic visualization chamber for watching DNA polymerase as it performs sequencing by synthesis. Assembly is done by stitching together individual reads.

NABsys is also using solid state, electrically addressable nanopore arrays. However, in contrast to Pacific Biosciences, the NABsys platform doesn’t use optics and isn’t watching the polymerase. Probes are hybridized to individual DNA molecules and the distance between them on each fragment is measured. Algorithms are used to reconstruct the genomic sequence.

More notes on each of the talks can be found on FriendFeed (ISB 10th International Symposium) and Healthcare Hashtags for Twitter (ISB2011P4).