Blavatnik Family Foundation and the New York Academy of Sciences
The 2015 Blavatnik Science Symposium

Posted October 27, 2015
Presented By
Overview
On August 5–6, 2015, the New York Academy of Sciences hosted the second annual Blavatnik Science Symposium, which showcased the work of winners and finalists of the Blavatnik Awards for Young Scientists. Founded by the Blavatnik Family Foundation in 2007 to recognize exceptional young scientists, the awards were initially a regional competition for postdocs and young scientists in New York, New Jersey, and Connecticut. The national competition, launched in 2014, now recognizes promising faculty scientists and engineers nationwide. Award winners and finalists receive unrestricted research funds and become Blavatnik Science Scholars, members of an interdisciplinary community of scientists who meet annually to learn about one another's research and to explore collaborations.
The symposium featured research updates from past and current winners and finalists from the national and regional awards programs. The speakers described their recent research activities in a symposium format designed to encourage questions and discussion among attendees. Several speakers also participated in panel discussions on topics such as artificial intelligence, entrepreneurship, and interdisciplinary partnerships in science.
Use the tabs above to find a meeting report and multimedia from this event.
Presentations available from:
David Blei, PhD (Columbia University)
Léon Bottou, PhD (Facebook AI Research)
Stephen Brohawn, PhD (The Rockefeller University)
Geoffrey Coates, PhD (Cornell University)
Matthew Disney, PhD (Scripps Research Institute, Florida)
Knut Drescher, PhD (Max Planck Institute for Terrestrial Microbiology, Germany)
Xiangfeng Duan, PhD (University of California, Los Angeles)
Eric Ford, PhD (Pennsylvania State University)
Alison Galvani, PhD (Yale University)
Markus Greiner, PhD (Harvard University)
Clément Hongler, PhD (École Polytechnique Fédérale de Lausanne, France)
Rob Knight, PhD (University of California, San Diego)
Nevan Krogan, PhD (University of California, San Francisco; Gladstone Institutes)
Laura Landweber, PhD (Princeton University; Columbia University)
Yann LeCun, PhD (Facebook AI Research; New York University)
Hakho Lee, PhD (Massachusetts General Hospital)
Hening Lin, PhD (Cornell University)
Michal Lipson, PhD (Columbia University)
Mary Kay Lobo, PhD (University of Maryland)
Yueh-Lin Loo, PhD (Princeton University)
Szabolcs Márka, PhD (Columbia University)
Luciano Marraffini, PhD (The Rockefeller University)
Ruslan Medzhitov, PhD (Yale University)
Evgeny Nudler, PhD (New York University)
Aydogan Ozcan, PhD (University of California, Los Angeles)
Jérémie Palacci, PhD (University of California, San Diego)
Jeremy Palmer, PhD (University of Houston)
Abhay Pasupathy, PhD (Columbia University)
Panteleimon Rompolas, PhD, MBA (University of Pennsylvania)
Jared Rutter, PhD (University of Utah)
Peng Yin, PhD (Harvard University)
Carl Zimmer, (The New York Times)
Yi Zuo, PhD (University of California, Santa Cruz)
Presented by
How to cite this eBriefing
The New York Academy of Sciences. The 2015 Blavatnik Science Symposium. Academy eBriefings. 2015. Available at: www.nyas.org/Blavatnik2015-eB
Resources
David Blei
Gopalan PK, Blei DM. Efficient discovery of overlapping communities in massive networks. Proc Natl Acad Sci U S A. 2013;110(36):14534-9.
Manning JR, Ranganath R, Norman KA, Blei DM. Topographic factor analysis: a Bayesian model for inferring brain networks from neural data. PLoS One. 2014;9(5):e94914.
Léon Bottou
Bottou L. From machine learning to machine reasoning. Mach Learn. 2014;94(2):133-49.
Oquab M, Bottou L, Laptev I, Sivic J. Learning and transferring mid-level image representations using convolutional neural networks. PROC CVPR IEEE. 2014;1717-24.
Stephen Brohawn
Brohawn SG, Su Z, MacKinnon R. Mechanosensitivity is mediated directly by the lipid membrane in TRAAK and TREK1 K+ channels. Proc Natl Acad Sci U S A. 2014;111(9):3614-9.
Geoffrey Coates
Hugar KM, Kostalik HA 4th, Coates GW. Imidazolium cations with exceptional alkaline stability: a systematic study of structure-stability relationships. J Am Chem Soc. 2015;137(27):8730-7.
Khurana R, Schaefer JL, Archer LA, Coates GW. Suppression of lithium dendrite growth using cross-linked polyethylene/polyethylene oxide electrolytes: a new approach for practical lithium-metal polymer batteries. J Am Chem Soc. 2014;136(20):7395-402.
Longo, JM, DiCiccio AM, Coates GW. Poly(propylene succinate): a new polymer stereocomplex. J Am Chem Soc. 2014;136(45):15897-900.
Matthew Disney
Paul DJ, Seedhouse SJ, Disney MD. Two-dimensional combinatorial screening and the RNA Privileged Space Predictor program efficiently identify aminoglycoside-RNA hairpin loop interactions. Nucleic Acids Res. 2009;37(17):5894-907.
Rzuczek S, Park H, Disney MD. A toxic RNA catalyzes the in cellulo synthesis of its own inhibitor. Angew Chem Int Ed Engl. 2014;6:53(41)10956-9.
Velagapudi SP, Gallo SM, Disney MD. Sequence-based design of bioactive small molecules that target precursor microRNAs. Nat Chem Biol. 2014;10(4):291-7.
Knut Drescher
Drescher K, Nadell CD, Stone HA, et al. Solutions to the public goods dilemma in bacterial biofilms. Curr Biol. 2014;24(1):50-5.
Xiangfeng Duan
Xu Y, Lin Z, Huang X, et al. Functionalized graphene hydrogel-based high performance supercapacitors. Adv Mater. 2013;25(40):5779-84.
Xu Y, Lin Z, Zhong X, et al. Holey graphene frameworks for highly efficient capacitive energy storage. Nat Commun. 2014;5:4554.
Eric Ford
Lissauer JJ, Fabrycky DC, Ford EB, et al. A closely packed system of low-mass, low-density planets transiting Kepler-11. Nature. 2011;470(7332):53-8.
Alison Galvani
Atkins KE, Shim E, Carroll S, et al. The cost-effectiveness of pentavalent rotavirus vaccination in England and Wales. Vaccine. 2012;30(48):6766-76.
Fitzpatrick MC, Hampson K, Cleaveland S, et al. Cost-effectiveness of canine vaccination to prevent human rabies in rural Tanzania. Ann Intern Med. 2014;160(2):91-100.
Pandey A, Atkins KE, Medlock J, et al. Strategies for containing Ebola in West Africa. Science. 2014;346(6212):991-5.
Markus Greiner
Bakr WS, Gillen JI, Peng A, et al. A quantum gas microscope for detecting single atoms in a Hubbard-regime optical lattice. Nature. 2009;462(7269):74-7.
Gillen JI, Bakr WS, Peng A, et al. Two-dimensional quantum gas in a hybrid surface trap. Phys Rev. 2009;A80,021602(R).
Clément Hongler
Chelkak D, Hongler C, Izyurov K. Conformal invariance of spin correlations in the planar Ising model. Prin Ann Math. 2014;181(3):1087-138.
Hongler C, Kytölä K. Ising interfaces and free boundary conditions. J Amer Math Soc. 2013;26:1107-89.
Hongler C, Smirnov S. The energy density in the planar Ising model. Acta Math. 2013; 211(2):191-225.
Rob Knight
Hsiao EY, McBride SW, Hsien S, et al. Microbiota modulate behavioral and physiological abnormalities associated with neurodevelopmental disorders. Cell. 2013;155(7):1451-63.
Integrative HMP (iHMP) Research Network Consortium. The Integrative Human Microbiome Project: dynamic analysis of microbiome-host omics profiles during periods of human health and disease. Cell Host Microbe. 2014;16(3):276-89.
Weingarden A, González A, Vázquez-Baeza Y, et al. Dynamic changes in short-and long-term bacterial composition following fecal microbiota transplantation for recurrent Colostridium difficile infection. Microbiome. 2015;3:10.
Nevan Krogan
Jäger S, Cimermancic P, Gulbahce N, et al. Global landscape of HIV-human protein complexes. Nature. 2011;481(7381):365-70.
Jäger S, Kim DY, Hultquist JF, et al. Vif hijacks CBF-β to degrade APOBEC3G and promote HIV-1 infection. Nature. 2011;481(7381):371-5.
Laura Landweber
Chen X, Bracht JR, Goldman AD, et al. The architecture of a scrambled genome reveals massive levels of genomic rearrangement during development. Cell. 2014;158(5):1187-98.
Fang W, Landweber LF. RNA-mediated genome rearrangement: hypotheses and evidence. Bioessays. 2013;35(2):84-7.
Swart EC, Bracht JR, Magrini V, et al. The Oxytricha trifallax macronuclear genome: a complex eukaryotic genome with 16,000 tiny chromosomes. PLoS Biol. 2013;11(1):e1001473.
Yann LeCun
LeCun Y, Bengio Y, Hinton G. Deep learning. Nature. 2015;521(7553):436-44. Review.
Hakho Lee
Haun JB, Castro CM, Wang R, et al. Micro-NMR for rapid molecular analysis of human tumor samples. Sci Transl Med. 2011;3(71):71ra16.
Lee H, Sun E, Ham D, Weissleder R. Chip-NMR biosensor for detection and molecular analysis of cells. Nat Med. 200;14(8):869-74.
Liong M, Hoang AN, Chung J, et al. Magnetic barcode assay for genetic detection of pathogens. Nat Commun. 2013;4:1752.
Hening Lin
Du J, Zhou Y, Su X, et al. Sirt5 is a NAD-dependent protein lysine and demalonylase and desuccinalyse. Science. 2011;334(6057):806-9.
Jiang H, Khan S, Wang Y, et al. SIRT6 regulates TNF-α secretion through hydrolysis of long-chain fatty acyl lysine. Nature. 2013;496(7443):110-3.
Michal Lipson
Cardenas J, Poitras CB, Robinson JT, et al. Low loss etchless silicon photonics waveguides. Opt Express. 2009;17(6):4752-7.
Liu A, Jones R, Liao L, et al. A high-speed silicon optical modulator based on a metal-oxide semiconductor capacitor. Nature. 2004;427(6975):615-8.
Xu Q, Schmidt B, Pradhan S, et al. Micrometre-scale silicon opto-electric modulator. Nature. 2005;435(7040):325-7.
Yang AH, Moore SD, Schmidt BS, et al. Optical manipulation of nanoparticles and biomolecules in sub-wavelength slot waveguides. Nature. 2009;457(7225):71-5.
Mary Kay Lobo
Chandra R, Francis TC, Konkalmatt P, et al. Opposing the role of Egr3 in nucleus accumbens cell subtypes in cocaine action. J Neurosci. 2015;35(20):7927-37.
Francis TC, Chandra R, Friend DM, et al. Nucleus accumbens medium spiny neuron subtypes mediate depression-related outcomes to social defeat stress. Biol Psychiatry. 2015;77(3):212-22.
Lobo MK, Covington HE 3rd, Chaudhury D, et al. Cell-type specific loss of BDNF signaling mimics optogenetic control of cocaine reward. Science. 2010;330(6002):385-90.
Yueh-Lin Loo
Hailey AK, Wang S-Y, Chen Y, et al. Quantifying the energy barriers and elucidating the charge transport mechanisms across interspherulite boundaries in solution-processed thin films. Adv Funct Mater. 2015. [Epub]
Lee SS, Tang SB, Smilgies DM, et al. Guiding crystallization around bends and sharp corners. Adv Mater. 2012;24(20):2692-8.
Szabolcs Márka
Mendes CS, Bartos I, Akay T, et al. Quantification of gait parameters in freely walking wild type and sensory deprived Drosophila melanogaster. Elife. 2013;2:e00231.
Mendes CS, Bartos I, Márka Z, et al. Quantification of gait parameters in freely walking rodents. BMC Biol. 2015;13:50.
Luciano Marraffini
Barrangou R, Fremaux C, Deveau H, et al. CRISPR provides acquired resistance against viruses in prokaryotes. Science. 2007;315(5819):1709-12.
Goldberg GW, Jiang W, Bikard D, et al. Conditional tolerance of temperate phages via transcription-dependent CRISPR-Cas targeting. Nature. 2014;514(7524):633-7.
Heler R, Samai P, Modell JW, et al. Cas9 specifies functional targets during CRISPR-Cas adaptation. Nature. 2015;519(7542):199-202.
Marraffini LA, Sontheimer EJ. CRISPR interference limits horizontal gene transfer in staphylococci by targeting DNA. Science. 2008;322(5909):1843-5.
Ruslan Medzhitov
Palm NW, Rosenstein RK, Yu S, et al. Bee venom phospholipase A2 induces a primary type 2 response that is dependent on the receptor ST2 and confers protective immunity. Immunity. 2013;39(5):976-85.
Florsheim E, Yu S, Bragatto I, et al. Integrated innate mechanisms involved in airway allergic inflammation to the serine protease subtilisin. J Immunol. 2015;194(10):4621-30.
Evgeny Nudler
Gusarov I, Gautier L, Smolentseva O, et al. Bacterial nitric oxide extends the lifespan of C. elegans. Cell. 2013;152(4):818-30.
Aydogan Ozcan
Greenbaum A, Zhang Y, Feizi A, et al. Wide-field computational imaging of pathology slides using lens-free on-chip microscopy. Sci Transl Med. 2014;6(267):267ra175.
Wei Q, Luo W, Chiang S, et al. Imaging and sizing of single DNA molecules on a mobile phone. ACS Nano. 2014;8(12):12725-33.
Jérémie Palacci
Palacci J, Sacanna S, Steinberg AP, et al. Living crystals of light-activated colloidal surfers. Science. 2013;339(6122):936-40.
Jeremy Palmer
Palmer JC, Martelli F, Liu Y, et al. Metastable liquid-liquid transition in a molecular model of water. Nature. 2014;510(7505):385-8.
Abhay Pasupathy
Goncher SJ, Zhao L, Pasupathy AN, Flynn GW. Substrate level control of the local doping in graphene. Nano Lett. 2013;13(4):1386-92.
Panteleimon Rompolas
Rompolas P, Deschene ER, Zito G. et al. Live imaging of stem cell and progeny behavior in physiological hair follicle regeneration. Nature. 2012;487(7408):496-9.
Rompolas P, Greco V. Stem cell dynamics in the hair follicle niche. Semin Cell Dev Biol. 2014;25-26:34-42.
Rompolas P, Mesa KR, Greco V. Spatial organization within a niche as a determinant of stem-cell fate. Nature. 2013;502(7472):513-8.
Jared Rutter
Bricker DK, Taylor EB, Schell JC, et al. A mitochondrial pyruvate carrier required for pyruvate uptake in yeast, Drosophila, and humans. Science. 2012;337(6090):96-100.
Schell JC, Olson KA, Jiang L, et al. A role for the mitochondrial pyruvate carrier as a repressor of the Warburg effect and colon cancer growth. Mol Cell. 2014;56(3):400-13.
Marin Soljačić
Hsu CW, Zhen B, Lee J, et al. Observation of trapped light within the radiation continuum. Nature. 2013;499(7457):188-91.
Lu L, Wang Z, Ye D, et al. TOPOLOGICAL MATTER: Experimental observation of Weyl points. Science. 2015;349(6248):622-4.
Shen Y, Ye D, Celanovic I, et al. Optical broadband angular selectivity. Science. 2014;343(6178):1499-501.
Peng Yin
Green AA, Silver PA, Collins JJ, Yin P. Toehold switches: de-novo-designed regulators of gene expression. Cell. 2014;159(4):925-39.
Jin Z, Sun W, Ke Y, et al. Metallized DNA lithography for encoding and transferring spatial information for graphene patterning. Nat Commun. 2013;4:1663.
Jungmann R, Avendaño MS, Woehrstein JB, et al. Multiplexed 3D cellular super-resolution imagine with DNA-PAINT and Exchange-PAINT. Nat Methods. 2014;11(3):313-8.
Sadowski JP, Calvert CR, Zhang DY, et al. Developmental self-assembly of a DNA tetrahedron. ACS Nano. 2014;8(4):3251-9.
Carl Zimmer
Zimmer C. Why do we have allergies? Mosaic; 7 April 2015.
Feng Zhang
Cong L, Ran FA, Cox D, et al. Multiplexed genome engineering using CRISPR/Cas systems. Science. 2013;339(6121):819-23.
Zhang F, Cong L, Lodato S, et al. Efficient construction of sequence-specific TAL effectors for modulating mammalian transcription. Nat Biotechnol. 2011;29(2):149-53.
Yi Zuo
Chen CC, Gilmore A, Zuo Y. Study motor skill learning by single-pellet reaching tasks in mice. J Vis Exp. 2014;(85);e51238.
Fu M, Yu X, Lu J, Zuo Y. Repetitive motor learning induces coordinated formation of clustered dendritic spines in vivo. Nature. 2012;483(7387):92-5.
Keynote Speakers
Michal Lipson, PhD
Columbia University
2010 Regional Award Winner – Faculty
profile | publications
Carl Zimmer
The New York Times
website
Speakers
David Blei, PhD
Columbia University
2013 Regional Award Winner – Faculty
profile | publications
Léon Bottou, PhD
Facebook AI Research
2007 Regional Award Winner – Faculty
profile | publications
Stephen Brohawn, PhD
The Rockefeller University
2014 Regional Award Finalist – Postdoc
profile | publications
Geoffrey Coates, PhD
Cornell University
2007 and 2008 Regional Award Finalist – Faculty
profile | publications
Matthew Disney, PhD
Scripps Research Institute, Florida
2015 National Award Finalist
website | publications
Knut Drescher, PhD
Max Planck Institute for Terrestrial Microbiology, Germany
2014 Regional Award Finalist – Postdoc
profile | publications
Xiangfeng Duan, PhD
University of California, Los Angeles
2015 National Award Finalist
website | publications
Eric Ford, PhD
Pennsylvania State University
2015 National Award Finalist
website | publications
Alison Galvani, PhD
Yale University
2012 Regional Award Winner – Faculty
profile | publications
Markus Greiner, PhD
Harvard University
2015 National Award Finalist
website | publications
Clément Hongler, PhD
École Polytechnique Fédérale de Lausanne, France
2014 Regional Award Winner – Postdoc
profile | publications
Rob Knight, PhD
University of California, San Diego
2014 and 2015 National Award Finalist
website | publications
Nevan Krogan, PhD
University of California, San Francisco; Gladstone Institutes
2015 National Award Finalist
website | publications
Laura Landweber, PhD
Princeton University; Columbia University
2008 Regional Award Winner – Faculty
profile | publications
Yann LeCun, PhD
Facebook AI Research; New York University
2014 and 2015 National Judge
website | publications
Hakho Lee, PhD
Massachusetts General Hospital
2015 National Award Finalist
website | publications
Hening Lin, PhD
Cornell University
2015 National Award Finalist
website | publications
Mary Kay Lobo, PhD
University of Maryland
2011 Regional Award Finalist – Postdoc
profile | publications
Yueh-Lin Loo, PhD
Princeton University
2015 National Award Finalist
website | publications
Szabolcs Márka, PhD
Columbia University
2011 Regional Award Winner – Faculty
profile | publications
Luciano Marraffini, PhD
The Rockefeller University
2015 National Award Finalist
website | publications
Ruslan Medzhitov, PhD
Yale University
2007 Regional Award Winner – Faculty
profile | publications
Evgeny Nudler, PhD
New York University
2010 Regional Award Winner – Faculty
profile | publications
Aydogan Ozcan, PhD
University of California, Los Angeles
2014 and 2015 National Award Finalist
website | publications
Jérémie Palacci, PhD
University of California, San Diego
2014 Regional Award Finalist – Postdoc
profile | publications
Jeremy Palmer, PhD
University of Houston
2014 Regional Award Winner – Postdoc
profile | publications
Abhay Pasupathy, PhD
Columbia University
2015 National Award Finalist
website | publications
Panteleimon Rompolas, MBA, PhD
University of Pennsylvania
2014 Regional Award Winner – Postdoc
profile | publications
Jared Rutter, PhD
University of Utah
2015 National Award Finalist
website | publications
Marin Soljačić, PhD
Massachusetts Institute of Technology
2014 National Award Winner – Faculty
profile | publications
Peng Yin, PhD
Harvard University
2014 and 2015 National Award Finalist
website | publications
Feng Zhang, PhD
Massachusetts Institute of Technology
2015 National Award Finalist
website | publications
Yi Zuo, PhD
University of California, Santa Cruz
2015 National Award Finalist
website | publications
Hallie Kapner
Hallie Kapner is a science writer based in Chappaqua, NY. She works with research universities and scientific organizations in the New York area, and has been writing about science for lay audiences and the media for more than 15 years. She has also written for the New York Academy of Sciences Magazine.
Keynote Speakers
Michal Lipson
Columbia University
Carl Zimmer
The New York Times
Speed-of-light computing
Keynote speaker Michal Lipson of Columbia University opened the symposium with a visual exercise for the audience. "Imagine the computer on your desk," she said. "If you were to open the case and turn off the cooling fan, you could cook an egg."
Power dissipation is a major limiting factor in efforts to improve computer performance. To meet demand for faster data propagation and better processing efficiency, manufacturers must contend with the component-melting heat generated in the process of moving data "half a centimeter of distance, from the computer's memory to the processor," Lipson said. At large data centers such as those owned by Google and Microsoft, the greatest proportion of energy usage goes to cooling the equipment. The problem, and the solution Lipson is researching, lies in the material used to move data that crucial half-centimeter. Today, copper wires conduct electrical energy; but tomorrow's computers will run cool, powered by light.
Silicon photonics aims to transform the power paradigm in computing by replacing metal interconnects with optics, swapping electrical signals for pulses of light. "Fiber is always cold, and light fundamentally doesn't consume power in its propagation. Optics is the only way to move a lot of data without heat," Lipson explained.
Using optical fibers to move data is not a new idea: large optical cables have been moving data over short and long distances for a generation. The innovation is to do so at the nanoscale, constructing nanoscale silicon waveguides with extremely high performance. "Using lithography, we can write a path for light on silicon wafers and sculpt its direction," she said. "If you send light into a channel where the walls have a high rate of refraction, as they do in silicon, it will follow naturally and stay there."
Understanding how to actively control light, rather than passively direct it through a design, became a major challenge. But breakthrough modulators and switches—active devices—can now control the flow and speed of light on a chip, allowing silicon photonics to flourish in computing. IBM and other companies have designed multiplexed silicon chips integrating optical and microelectric circuits.
One technique, Lipson acknowledged, sounds more like fiction than fact: invisibility cloaking. The human eye perceives an object when light hits the object, scatters, and returns to the eye at various angles. But if light arrives at the eye in straight lines, no object is perceived. Lipson's team has successfully deployed silicon-based materials that can steer reflected light into a straight path to a viewer, rendering objects up to several millimeters in size invisible.
Another application pairs silicon photonic waveguides with flexible microelectronics to create a new generation of neural probe capable of directing light into the brain to stimulate function.

A model of a multiplexed chip integrating optical and microelectric circuits. (Image courtesy of Michal Lipson)
Cross talk—Telling stories about science
"To the immunologists in the audience, this may be a delightful, wonderful, illuminating thing—but to a lot of people it's incredibly terrifying. They want to run away screaming." Carl Zimmer was not visiting a lab or confronting specimens of an infectious virus—he saves those encounters for his reporting for the New York Times Science Section and other science publications. Zimmer was addressing symposium attendees, and projected behind him was a complex model of a late-phase allergic response. The audience laughed.

Complex mechanisms of a late-phase allergic response. (Image presented by Carl Zimmer courtesy of Hodge and Sayers. Allergy – Encyclopedia of Life Sciences)
So began Zimmer's keynote address, during which he discussed his process—refined over more than two decades of science writing—for turning the extraordinary work of scientists into highly readable stories for the public.
The allergy graphic was no accident. The story it relates to profiles the work of a 2007 Blavatnik regional faculty winner, Ruslan Medzhitov, who earlier this year published a controversial hypothesis of allergic response. Rather than an immunological error, it suggested, allergies could be a sophisticated defense against toxins. Intrigued by the research, Zimmer contacted Medzhitov to pursue a story. Zimmer's comments explored the makings of, and how to tell, a good story.
"I love writing stories about things in our experience that seem familiar—we think we know what they are, but science can help to show us that we didn't ... and in fact they're more interesting than we thought," he said of the need to choose broadly resonant topics. Writers should also remember that people are behind science; communicating their humanity can add a rich dimension to science stories. "You're writing about people, not just circles and arrows. You're writing about people who are thinking about circles and arrows," he said.
At the core of science writing is science itself, the language of which poses considerable challenges for writers—and readers. "I know I will not be quoting this in the article," Zimmer joked, after reading aloud a paragraph from Medzhitov's paper outlining his hypothesis. Zimmer spoke about the ongoing struggle to convey complex concepts in simple ways that are "understandable to the public but doing justice to the science."
Communicating about critical issues to a skeptical and often science-naïve public is a task Zimmer does not take lightly. He discussed how misunderstandings can have devastating effects on public health, citing communication about the safety of vaccines as a prominent example.
Zimmer lamented the decline of science content in newspapers but touted the rise of websites for science journalism and opportunities for scientists to self-publish through blogs. Despite the difficulties involved in science writing, Zimmer effusively encouraged scientists to share their work in their own words. "Part of being a scientist is being a public scientist; that's just something I believe," he said.

A humorous depiction of the science communications process, from bench to headline. (Image courtesy of Carl Zimmer)
Speakers
Matthew Disney
Scripps Research Institute, Florida
Yi Zuo
University of California, Santa Cruz
Hakho Lee
Massachusetts General Hospital
Feng Zhang
Massachusetts Institute of Technology
Nevan Krogan
University of California, San Francisco
Luciano Marraffini
The Rockefeller University
Jared Rutter
University of Utah
Sequence-based design of precision medicines targeting RNA
Matthew Disney of Scripps Florida opened the first session of talks by 2015 national finalists. He began by noting that advances in genome sequencing have identified disease-associated genes and proteins as targets for precision medicine.
Many drugs target protein products, but as new sequencing techniques have shown, only 1% of the human genome encodes proteins, while more than 80% encodes RNAs. Non-coding RNAs—which are not translated into proteins—are implicated in cancer, heart disease, muscular dystrophy, and many other diseases. Thus, RNAs too are targets for small-molecule therapeutics.
RNAs fold to form predictable structures of matching base-pair sequences, with unique loops and bulges. Computational analysis allows researchers to predict the structure of an RNA from its sequence, and then compile databases of RNA motifs. Disney and his colleagues aim to discover which small molecules bind these RNA structures. "We can take a lead drug and expose it to hundreds of thousands of RNA motifs and see which ones the molecules like best," Disney said. He used a lock-and-key analogy to describe testing thousands of small molecules against the RNA library. After identifying drugs that bind RNAs, the next step is testing which interactions are important in disease.
Disney described his work to identify small molecules that bind cancer-associated microRNAs, with the aims of inhibiting pathways that drive cancer and promoting apoptosis in cancer cells.

Unique loop structures in disease-associated RNAs are targets for small-molecule therapeutics. (Image courtesy of Matthew Disney)
Disease diagnostics in your hand
Hakho Lee of Massachusetts General Hospital described his work to develop a new class of biosensors for portable, affordable medical diagnostic devices. Point-of-care devices are in high demand, especially in developing countries. Despite tremendous technological progress, preventable diseases kill millions of people each year. Tuberculosis, malaria, and cancer are among the diseases for which low-cost diagnostic devices are particularly useful.
Lee's devices combine nanoparticles, which amplify signals at the cellular level, with microfluidics to enable high-throughput assays and with microelectronics to facilitate portability. Showcasing prototype point-of-care diagnostic devices compatible with smartphones, Lee explained how a postage-stamp-sized computer chip can be customized so that the system works for various diseases. He designed a miniaturized nuclear magnetic resonance (NMR) system that can detect and profile individual cancer cells with up to 96% accuracy. The same system can also detect tuberculosis, returning results within about 30 minutes.

Magnetic nanoparticles facilitate disease diagnosis in a novel point-of-care MicroNMR device that can detect tuberculosis and cancer. (Image courtesy of Hakho Lee)
In search of motor memories
Yi Zuo of the University of California, Santa Cruz, discussed one of the most complex, mysterious machines ever studied—the human brain. Neural networks constantly change in response to the environment, with feedback mechanisms guiding activity and behavior. Scientists only poorly understand how neural networks change and how information is stored in synapses.
Zuo studies motor memory, which consolidates information about physical tasks performed repetitively. Using transgenic mice with fluorescent tags on select motor neurons, her team imaged structural changes in the synapses over time.
As an animal attempts a new task, new synaptic connections form rapidly; when the task is mastered, synapse formation ends. "Making new connections is tightly associated with improvements in behavior," Zuo said. Synapses are selectively stabilized during the memory consolidation phase, and these synapses remain functional over the long term. "Once the network is set up, [mice] don't need to make new connections, even if we ask the mice to perform the task months after the initial learning," she said.
Additional studies by Zuo's team and others showed that when animals learn related tasks, the new activities are encoded by a different set of synapses clustered near the spines formed during the initial training; thus each motor task has its own memory network, with similar tasks consolidated in close proximity.
Bacterial immunity and genome editing
Luciano Marraffini of Rockefeller University gave an overview of the CRISPR/Cas system, the adaptive immune system of bacteria. He noted that all organisms are subject to viral infection, and that each has some form of immune system. The mechanism by which bacteria develop resistance to bacteriophages (phages) and plasmids is a fairly recent discovery.
Clustered regularly interspaced short palindromic repeats (CRISPRs) are repetitive DNA sequences on the genome of bacteria, accompanied by CRISPR-associated (Cas) genes. CRISPR sequences are separated by spacer sequences that match sequences of phage DNA. These spacer sequences form the basis of CRISPR immunity. "If a bacterium possesses a spacer that matches a viral genome, then that virus cannot infect that bacterium," Marraffini explained.
Each spacer sequence in bacterial DNA is transcribed as RNA, which couples with Cas proteins to form a complex that scans for phage DNA and other foreign DNA. If the bacterium has previously encountered a phage, the complex matches the RNA guide with the phage DNA sequence. Cas proteins are nucleases, which quickly cleave the viral DNA and prevent its replication.
One of the most important findings from Marraffini's research is the discovery that CRISPR immunity targets viral DNA, not viral RNA (which would also have accounted for immunity). "This opened possibilities for applications well beyond the microbial world," he said.
Neuroscientist Feng Zhang of the Massachusetts Institute of Technology talked about those possibilities, which include the ability to quickly and precisely edit DNA in any species, including humans. He focused on the CRISPR/Cas9 system—a relatively simple CRISPR/Cas system that relies on one enzyme (Cas9)—and its applications in brain disorders.
Despite advances in optogenetics and neuroimaging, the mechanisms of neurological diseases are not well understood. However, if the goal is to develop therapeutics, "everything boils down to the genome," Zhang said. "So the question becomes, how do we target that?" Although techniques for genome editing had been developed before the discovery of CRISPR/Cas, all were costly and labor-intensive.
The "elegant microbial adaptive immune system" piqued Zhang's interest in 2010, and he began to explore whether the system could be adapted to cleave DNA in a eukaryotic cell. He and his colleagues first published on CRISPR/Cas9 in 2013, after several years of work to adapt the microbial system to function in mammalian cells. Guided by two strands of RNA, CRISPR/Cas9 can bind to a precise location on human DNA and deploy Cas9 to snip the desired section to either replace a defective gene or allow the DNA to repair itself.
Zhang discussed the extraordinary media response to the discovery of CRISPR/Cas and its applications, and discussed his lab's efforts to improve the technology and to create a community for sharing information. He concluded with an example use for the CRISPR/Cas9 system, discussing mouse experiments in which his lab modified a gene associated with autism spectrum disorders.
A systems approach to studying HIV
Nevan Krogan of the University of California, San Francisco, described a systems biology approach to mapping human–viral protein interactions in human cells. HIV-1 has a small genome, and Krogan's team expressed each HIV protein in various cell types and then used mass spectrometry to identify the associated human proteins.
One HIV protein, Vif, had been previously implicated in HIV pathogenesis. Vif compromises the immune response to HIV by hijacking an ubiquitin ligase complex and targeting for degradation the restriction factor APOBEC3G, which blocks viral replication.
Yet "nothing was known about the structure of this protein," Krogan said. Researchers had been unable to reconstitute it within the complex in vitro. "But if you had the structure, hopefully that would lead to more intelligent drug design."
Krogan confirmed his hypothesis that another protein is involved in the ubiquitin ligase complex, the transcription factor CBFB. Addition of CBFB to the other known components of the complex reconstituted the complex in vitro. If CBFB is inhibited, the team found, Vif cannot hijack APOBEC3G—and HIV infection is nearly impossible. The researchers also characterized the structure of Vif. Krogan and others are now studying therapeutic possibilities for interrupting the association between Vif and CBFB, or other proteins within the ubiquitin ligase complex, to block HIV infection.

A map of human–viral protein interactions was the first step in discovering the association between Vif and the transcription factor CBFB in HIV infectivity. (Image courtesy of Nevan Krogan)
Mitochondria, metabolism, and cancer
Jared Rutter of the University of Utah focused on mitochondria, which produce adenosine triphosphate (ATP) to power cell functions. His work challenges our understanding of the metabolic network in cells, suggesting that rather than passively providing the energy and building blocks cells need to grow and divide, the network also actively affects cell behavior and cell fate.
Roughly one third of the human mitochondrial proteome is uncharacterized. "We know the proteins are there, but we have no idea what they're doing ... and many of these proteins are found across all eukaryotes," Rutter said.
His team discovered that mitochondrial pyruvate carrier (MPC)—a protein that plays a role in cancer cell fate—is responsible for allowing pyruvate, a product of glycolysis, to enter mitochondria and undergo oxidation in the process of generating ATP. Cancer cells do not always use pyruvate in this way, however, and often instead convert it into building blocks for cell growth and proliferation—a shift known as the Warburg effect.
Rutter tested the hypothesis that lack of pyruvate oxidation in cancer is caused by lost MPC function. Almost all solid tumors show decreased expression or mutation of this transporter. When tumor cells are forced to express a functional form of MPC, and hence to use pyruvate to generate ATP, there are "devastating consequences on these cells' ability to form tumors," Rutter said. All markers of stemness—crucial in tumor formation and metastasis—decrease when normal MPC expression is restored.
Speakers
Clément Hongler
École Polytechnique Fédérale de Lausanne, France
Knut Drescher
Max Planck Institute for Terrestrial Microbiology, Germany
Jeremy Palmer
University of Houston
Jérémie Palacci
University of California, San Diego
Panteleimon Rompolas
University of Pennsylvania
Stephen Brohawn
The Rockefeller University
Typical material, unusual behavior
Jeremy Palmer of the University of Houston gave the first talk by 2015 regional winners and finalists, exploring the surprising properties of a substance rarely considered exotic—water. This everyday liquid presents challenging questions that sophisticated modeling and simulation technologies can help answer.
Palmer described floating ice cubes as a familiar example of water's atypical behavior: most liquids become denser as they cool but ice crystals have a lower density than liquid water. Peculiar characteristics also emerge when water is supercooled below its melting point without forming a solid.
Supercooled water exhibits dramatic fluctuations in energy and density, particularly at temperatures around −45°C (228 K). This state is difficult to access experimentally. At even lower temperatures, water assumes a low- or high-density amorphous solid state, and scientists theorized that the fluctuations at temperatures around −45°C could be caused by phase transitions between low- and high-density liquid states. No studies have succeeded in accessing this temperature zone to explore this hypothesis.
Palmer used computer simulations to model the behavior of water in this temperature zone. His group was first to show that molecular models of water can exhibit its unusual behaviors when simulating phase transitions between two phases that are not thermodynamically stable.

Scientists have theorized about the properties of supercooled water at difficult-to-access temperatures above the crystallization threshold, positing in 1992 the existence of low- and high-density liquid states now confirmed by Palmer's computational models. (Image courtesy of Jeremy Palmer)
Bringing biological order to synthetic materials
Jérémie Palacci of the University of California, San Diego, opened with a review of equilibrium systems, illustrating with the example of drop of dye in a glass of water, which will quickly homogenize. Non-equilibrium (organized) states, and emergent properties, are possible when energy is added to a system. Most systems in nature are non-equilibrium states.
Experiments with colloids demonstrate how non-living systems can mimic the characteristics of living systems—including propulsion, interaction, and self-organization—when in a state of non-equilibrium.
He showed a video of gold-platinum nanorods in a hydrogen peroxide solution studded with nonreactive spheres. The rods and the solution create a small "battery," and the rods begin moving around the spheres. "If I didn't tell you it was a rod, you'd think it was a bacterium, and that it wants to do this," Palacci said. He explained that such experiments may pave the way for materials that can self-assemble or repair. "For a physicist, this is a good first step, but for a biologist, it's nothing compared to what cells can do," he said.
Discrete structures and emergent symmetries
Clément Hongler of the École Polytechnique Fédérale de Lausanne showed the audience two images—one disorganized, the other a visually complex work by the artist M.C. Escher (shown below). "The goal of my talk is to explain the connection between these two pictures, and what we can learn from these connections," Hongler said. He introduced lattice models, which describe the large-scale world made up of small entities (atoms or small molecules), with positive or negative spin, arranged on a grid.
The Critical Ising lattice model shows how each spin influences its neighboring spins, and how far the influence propagates. It has been known for decades that the strength of the local influence determines whether the system will proceed to global alignment or to global disorder. There is a critical point at which the system remains in flux. Quantum field theory is useful for understanding how spins affect each other at long ranges, if you know the conformal symmetries of a model.
In the Escher piece, the viewer can "map any fish onto any other fish," Hongler explained. This type of symmetry—conformal symmetry—is "extremely strong," he said. Conformal symmetry reveals the correlations among each of the small entities in the system.
Hongler illustrated his theorem, which proved that "the correlations of the Ising model are given in terms of conformal field theory correlations." Thus, "if you want to compute the correlation of the spins, you can just look at the picture and ask, 'what's the size of the fish here and the fish there? How many fish do you need to count to get from point a to point b?' We can use those numbers to create a formula for understanding the correlation between spins."

The Critical Ising model contains conformal field theory precursors. Correlations between spins can be computed by understanding the symmetries of the image. (Image courtesy of Clément Hongler)
Live-imaging skin regeneration
Skin cells have a short life span and are continuously replaced, but the activity of resident stem cells in the skin is not well understood. Panteleimon Rompolas of the University of Pennsylvania created a system to visualize stem-cell activity in live mice, which he demonstrated with two short videos of epidermal and hair-follicle stem cells in action.
The team tracked skin-cell regeneration by labeling these stem cell populations. Rompolas reported some of the more surprising results, such as the finding that a cell's location in the stem cell niche determines whether it is likely to differentiate or to remain uncommitted.
Experiments on hair-follicle cells show that cell regeneration stalls when one group of cells—in the mesenchymal niche—are ablated, suggesting that these cells provide instructions to those stem cells which directly contribute to hair growth. Conversely, if hair-follicle cells themselves are ablated, cell regeneration is unaffected. "Epithelial stem cells come into the same position and learn to behave as resident hair-follicle cells," Rompolas said. "Stem cells that normally do one thing can learn new tasks under certain circumstances."
Rompolas explained that his team is investigating the mechanisms of this plasticity among stem cells, which may lead to improved skin grafting techniques for burn or injury victims.
How bacteria solve the public goods dilemma
Knut Drescher of the Max Planck Institute for Terrestrial Microbiology explained how the bacterial species Vibrio cholerae solves the public goods dilemma—a social quandary that asks why individuals contribute to a communal asset from which all members stand to benefit, and which would not exist if none contributed, but which would exist without the contribution of any particular individual. Among V. cholerae, the asset is extracellular enzymes secreted to aid digestion of chitin. These bacteria live in dense, surface-associated communities called biofilms, and the extracellular enzymes, which liberate sugars from chitin, benefit bacteria that secrete them as well as bacteria with mutations that impair this capability.
The nonproducing "social cheaters" expend no metabolic energy contributing to the public good, and should theoretically be free to expend more energy on growth. If that were true, however, the ability to secrete these crucial extracellular enzymes would be evolutionarily lost. "When you take an environmental isolate of a bacteria and sequence it, you find that almost all isolates are capable of producing these public goods. So cheaters, in the natural world, are actually deselected for," Dresher said.
To test how these bacterial communities solve the dilemma and maintain the dominance of contributors, Drescher created an experimental environment mimicking natural conditions. He found that some bacterial strains possessed mutations that result in production of very thick biofilms. When these "hypersecretors" are present, the producers will always overcome the cheaters in a bacterial community. If the biofilm layer is 5-cells rather than 1-cell thick, much less liberated sugar escapes the biofilm, and a public good is converted into an asset for active producers. When there is a thin biofilm, nature also cheats the cheaters if the nutrients available from chitin digestion are washed away. In that case too, the social cheaters are denied access and the producers triumph.
Cell mechanisms of sensation
Stephen Brohawn of Rockefeller University focused on the cellular mechanisms that allow for sensation of mechanical forces. Cells with force-sensing (mechanosensitive) ion channels convert mechanical forces to electrical signals, producing a spiking electrical signal when prodded. This response to mechanical force accounts for perception of touch, hearing, balance, and pain. Brohawn's presentation focused on the TRAAK ion channel in neurons, which dampens pain sensation.
Brohawn's team found that TRAAK's mechanism is unique. The channel has three openings—one facing the outside of the cell, one facing inside, and one facing the membrane. The channel closes when a lipid molecule enters the side opening and binds within the channel to block ion flow across the membrane. In the open position, the channel expands to seal this side opening, preventing lipid entry and allowing free flow of ions.
TRAAK channels open in response to membrane stretching. "Imagine making a mark on a balloon and watching the dot stretch as you inflate the balloon," Brohawn explained. "The wider conformation is favored when the membrane is stretched." Ion flow out of the cell quiets neurons and dampens pain sensation. It may be possible to manipulate these channels pharmacologically to manage pain.

A lipid prevents ion flow through a TRAAK ion channel in the closed position. Membrane tension opens the channel. (Image courtesy of Stephen Brohawn)
Speakers
Ruslan Medzhitov
Yale University
Szabolcs Márka
Columbia University
Evgeny Nudler
New York University
Alison Galvani
Yale University
David Blei
Columbia University
Léon Bottou
Facebook AI Research
Yann LeCun
Facebook AI Research; New York University
Career reinvention, second careers, and cross-disciplinary research programs
The first day of the symposium closed with a panel discussion among four regional faculty winners from previous years. Carl Zimmer, New York Times science columnist, moderated a conversation about interdisciplinary projects that stretched the panelists' skills, challenged their understanding, and led to unexpected outcomes. Each panelist began with a brief introduction.
For Ruslan Medzhitov of Yale University, initial work in immunobiology led to an interest in evolutionary medicine, an emerging field that examines the evolutionary bases of human disease. It aims to uncover not proximal causes of disease but evolutionary explanations of disease processes. "It's the 'why' questions, not the 'how' or 'what' questions," Medzhitov said.
Astrophysicist Szabolcs Márka of Columbia University made an unexpected leap from studying black holes to studying malarial mosquitoes. Márka explained that his approach has yielded promising schemes for controlling mosquitoes, including an infrared mosquito "net" that repels the insects with no harm to humans.
Evgeny Nudler of New York University studies how commensal bacteria affect host physiology, with a focus on aging. Using Caenorhabditis elegans as a model, Nudler investigated why the worm's life expectancy doubles when its commensal bacteria, Escherichia coli, is replaced by Bacillus subtilis, which produces nitric oxide and other metabolites that increase the worm's resistance to environmental stressors.
Alison Galvani of Yale University uses mathematical modeling techniques to inform public policy decisions. She has worked on strategies to reduce Ebola virus transmission and to improve contact tracing in Liberia. She has also used the techniques to make recommendations for rotavirus vaccine policy in the UK and for canine vaccination as a method to prevent human rabies infection in Tanzania.
Zimmer and the panelists engaged in a lively discussion about curiosity and risk taking, as well as the value of intuition over advice. Early in his career, Nudler told the audience, advisers warned him against pursuing research outside of his area of focus at the time—the mechanics of transcription. "If I had followed their advice, I'd be broke," he said, explaining how a diverse portfolio of work has not only enriched his research outcomes but provided steady funding options.
Márka and other panelists expressed the usefulness of unrestricted funding, such as a Blavatnik Award, in easing anxiety associated with new academic pursuits. "I didn't switch fields, I enlarged my field—and we are all proving that you can do it," Márka said.
Several questions concerned the differing jargon and styles of work and communication in different fields. Zimmer asked whether a scientist might feel "like an undergraduate student" when beginning an interdisciplinary collaboration. Nudler replied that building knowledge in a new area is a process, and that the novelty is more exciting than unnerving. Medzhitov echoed this sentiment, saying that new entrants into a field can challenge existing mindsets.
Panelists and audience members discussed the value of gatherings like the symposium, where collaboration among fields is celebrated and attendees find previously unknown commonalities. As Medzhitov said, "Everyone makes a choice—you can stick with your comfort zone or move out of it."

Galvani's team designed a mobile app to transmit real-time contact tracing data from Ebola-affected regions of West Africa. (Image courtesy of Alison Galvani)
Artificial intelligence and machine learning—Can computers be truly intelligent?
The second day of the symposium began with a panel discussion on artificial intelligence. David Blei of Columbia University opened with an overview of the goals of modern machine learning. Machine learning analysis finds patterns in large data sets, such as human genome data, fMRI measurements, and computer networks. Machine learning is historically about finding patterns and making predictions, but with the advent of technologies that produce massive data sets, the field is changing. A probabilistic approach is now more common. Scientists and statisticians collaborate to form hypotheses, analyze data to test the hypotheses, detect patterns, and apply the resulting models to solve problems.
Léon Bottou of Facebook AI Research explained that the initial assumptions about the field—that computers could replicate intelligence—fell flat for a simple reason: early computers "had to be programmed ... they only knew how to do what you wanted them to do." He showed a series of typographical variants on the numeral 4, explaining that rules cannot define all the shapes that humans recognize as representing that numeral. The same is true of other visual objects. By contrast, human brains are not programmed, but educated, and learn by example. The field of artificial intelligence seeks to give computers this capability.
Deep learning attempts to mimic the human ability to learn by example, explained Yann LeCun of Facebook AI Research and New York University. He described the massive computational power required to teach a machine even a single distinction, such as that of an airplane from a car. This approach is called supervised learning. Teaching computers to classify information and make distinctions has many familiar applications in speech, facial, and visual recognition software used in smartphones, autonomous robots, and autonomous vehicles. LeCun is studying convolutional nets, which have allowed automated visual classification systems to recognize even very similar activities—such as ice skating, figure skating, and speed skating—as distinct.
The audience asked questions about unsupervised machine learning and about whether machines can exhibit common sense. LeCun noted that even the simplest acts of prediction still evade computers. Audience members debated what would constitute intelligence in a computer, and Blei pointed to intelligence as our abilities to learn multiple tasks simultaneously and to generalize skills. A computer learning to play a video game does not show intelligence, LeCun argued. "If the computer uses those skills to play a different game ... that's intelligence," he said.

Machine learning can find patterns in huge data sets. This slide shows the most common words in diplomatic cables from the 1970s. (Image courtesy of David Blei)
Speakers
Eric Ford
Pennsylvania State University
Xiangfeng Duan
University of California, Los Angeles
Abhay Pasupathy
Columbia University
Markus Greiner
Harvard University
Yueh-Lin (Lynn) Loo
Princeton University
Peng Yin
Harvard University
Hening Lin
Cornell University
The search for exoplanets and our place in the universe
Eric Ford of Pennsylvania State University opened a second set of presentations by 2015 national finalists with a talk about space exploration. Progress in discovering exoplanets was once measured in centuries, but now planets and planetary systems are discovered regularly, with data from NASA's Kepler Mission.
Kepler has mostly identified planets larger than Earth but smaller than Neptune, a range for which there exists no analogue in our solar system. And while it is possible to glean information about planetary size and orbital period from Kepler data, little is known about the composition of these exoplanets. "We don't know if they're mostly rock or mostly gas. ... We're trying to understand this new class of planet to figure out what is potentially the most common type of planet in the galaxy," Ford said.
Of particular interest is the potential for life on exoplanets. Despite uncertainties about the atmospheric conditions on individual planets, it is possible to determine which ones fall into the "habitable zone," with temperatures favorable to liquid water. There are a wide variety of configurations of planetary systems in our galaxy; most planetary systems differ from our solar system. Some systems have tightly spaced planets; others have large but low-mass planets that challenge conventional ideas about how planets form around a central star.
Graphene will power the super battery of the future
Xiangfeng Duan of the University of California, Los Angeles, presented a brief history of semiconductors, noting that a modern smartphone packs more computing power than a room full of computers from the 1970s. Demand for portable computing spurred the creation of smaller, denser, and faster electronics for these devices, while larger, lighter, and lower-performance electronics dominate information-display technologies, such as massive display screens in urban centers and sports venues.
There is no middle ground—a high-performance semiconductor material for both small and large applications does not exist. Duan discussed how 2D materials—crystalline materials consisting of a single layer of atoms—could serve this function. The most promising of these materials is graphene, a single atomic layer of graphite. Graphene is light and strong and has superior electronic capabilities: electrons travel up to 100 times faster in graphene than in silicon.

2D materials offer capability to combine and control properties at the atomic scale, and enable devices with unprecedented performance and unique functions. (Image courtesy of Xiangfeng Duan)
Duan is looking for ways to control graphene at the atomic scale, specifically for applications in energy storage. "Mobile power supply is the bottleneck holding back the evolution of new mobile devices and electric cars," Duan explained. He described his lab's experiments to devise a 3D graphene framework for use in supercapacitor electrodes. The energy and power density in such a "SuperBattery" would be increased 10-fold over today's supercapacitors, and the battery would have 100-fold higher power than a traditional battery.
Developments in materials science at the atomic scale
Abhay Pasupathy of Columbia University noted that human progress has relied on mastery of materials—first bronze and iron, and later copper, steel, and silicon. Today we are poised to enter the age of "quantum materials"; that is, we must understand the quantum nature of materials in new technologies to design practical applications.
Indeed, using the materials that power modern life, including those in metal–oxide–semiconductor field-effect transistor (MOSFET) chips, is not like using a material such as wood—"If [wood] falls on your head, you don't need quantum mechanics to know that's trouble," Pasupathy joked. We need to know how the electrons in these materials behave.
The wavelengths of electrons are much too short to be studied using traditional microscopy, so Pasupathy and his colleagues built custom instruments to study electron behavior and quantum properties. The team produced images of various phenomena using nitrogen-doped graphene, which has potential for use in semiconductors and sensors, and sodium iron arsenide, a semiconducting material that displays an unusual arrangement of electrons.
Specialized tools for studying quantum mechanics are also useful for designing applications for electronic devices. Pasupathy and his team designed a veselago lens that may outperform traditional silicon MOSFETs.
Markus Greiner of Harvard University discussed a new method for modeling the quantum properties of materials. By cooling atoms to temperatures barely above absolute zero, trapping them in an optical lattice, and imaging them with a quantum gas microscope, Greiner and his team observed individual atoms transitioning between different sites on the lattice, governed by quantum mechanics, with behavior mirroring that of electrons. This ability to simulate the behavior of electrons with atoms will advance understanding of the quantum properties of magnetic materials, semiconductors, and other materials.
Structural heterogeneities in plastic electronics
Lynn Loo of Princeton University gave the audience an update on advances in plastic electronics. Electrically conductive plastics are used in pressure sensors, solar panels, light-emitting diodes, and many other technologies. Loo explained that organic semiconductors and conducting polymers have particularly useful properties and chemical versatility. "Using simple chemistry, we can drastically change their electrical properties, mechanical properties, and optoelectronic properties," she said.
Small chemical changes enable a wide range of applications for these materials. Organic LEDs can be tweaked to emit different colors of light or to absorb light for use in a photovoltaic cell. Electrically conductive plastics and organic semiconductors can also be made into soluble inks and other formulations to cover large areas, as in polymer solar cells that can be tailored to cover surfaces of any shape.
Unlike inorganic semiconductors, organic semiconductors require that electrical charge hop from one molecule to another. Thus, charge transport depends on the molecular organization of the substrate. Loo's lab is working to understand how molecular organization affects charge transport. The team has discovered that different substrates promote different rates of crystallization, and has found success guiding crystallization in specific patterns to optimize charge transport.

Plastic electronics are lightweight, flexible, and tunable, and can be used in a wide range of applications. (Image courtesy of Lynn Loo)
Molecular programming with DNA
Peng Yin of Harvard University works with molecular structures called DNA bricks, which he likened to toys such as building blocks. Complementary segments of DNA bind together, and by "programming" or directing these segments, it is possible to coax DNA to assemble into user-defined 2D and 3D nanostructures.
DNA bricks can be used to control molecular-scale systems. Yin is exploiting the programmability of DNA to create molecular circuits in living cells. He explained that by integrating DNA bricks with gold, it is possible to program the morphology of inorganic materials, including graphene. He has used molecular programming to direct gene expression in bacteria, using an RNA-based "switch" to regulate genes.
Yin has also devised an inexpensive imaging technique called DNA-PAINT for visualizing cellular processes. It produces super-high-resolution images of synthetic DNA structures that are fluorescently tagged and transiently bind to complementary target structures, producing a "blink" upon detecting the target. The method could reveal cell processes in unprecedented detail.

Fluorescent DNA tags transiently bind to marker tags on a synthetic DNA structure. (Image courtesy of Peng Yin)
Demystifying orphan enzymes
Hening Lin of Cornell University continued the conversation about DNA, discussing two challenges of the postgenomic age: the unknown biochemical functions of many human proteins, and the mechanisms of posttranslational modifications, or PTMs, which diversify the proteome and play important roles in regulating cell processes.
Orphan enzymes—proteins classified as enzymes based on homology but without any known enzymatic activity—are thought to control PTMs. Lin's lab focuses on a class of orphan enzymes called sirtuins. There are 7 human sirtuins, all of which are presumed to catalyze removal of acetyl groups from proteins. Lin's research showed that just 3 sirtuins functioned this way. The others had very weak activity.
Further experiments showed that the weak sirtuins removed malonyl and succinyl groups, rather than acetyl groups, representing a new form of protein modification that has been implicated in heart metabolism and function and in cancer.
Speakers
Aydogan Ozcan
University of California, Los Angeles
Rob Knight
University of California, San Diego
Geoffrey Coates
Cornell University
Yueh-Lin (Lynn) Loo
Princeton University
Szabolcs Márka
Columbia University
Mary Kay Lobo
University of Maryland
Laura Landweber
Princeton University; Columbia University
Marin Soljačić
Massachusetts Institute of Technology
Commercialization and entrepreneurship
Three 2015 national finalists, Rob Knight, Aydogan Ozcan, and Lynn Loo, joined previous faculty finalist Geoffrey Coates for a panel discussion on commercialization and entrepreneurship opportunities.
"We've heard about life-saving health research today, but I just make plastic," joked Geoffrey Coates, who began the panel by describing the ubiquity and importance of modern plastics, as well as their drawbacks, including fossil fuel use and issues of biodegradability. Coates developed a proprietary set of catalysts that can produce plastics containing up to 50% sequestered CO2. The technology underlies Novomer, the company he cofounded in 2004.
Rob Knight took the podium next to explain how his company, Biota, uses microbial analysis to maximize production of new oil wells. Sampling and sequencing the unique microbial communities in oil wells can inform decisions about where to pursue drilling.
Aydogan Ozcan's company, Holomic, develops advanced diagnostics on a common platform: a smartphone. Ozcan devised a lens-free system that couples complementary metal–oxide–semiconductor (CMOS) chips with custom diagnostic software that can detect cancer, HIV, HSV, measles, and many other diseases. Designed for use in any setting, the devices communicate sample data to a server for interpretation, obviating the need for trained lab personnel on-site.
Lynn Loo discussed her experiences serving as director of the E-ffiliates Partnership program at Princeton University's Andlinger Center for Energy and the Environment. The program fosters collaborations and technology transfer between Princeton students and industries in the energy and environmental sectors.
The panelists touched on topics such as their impetuses for starting a business and struggles to balance an academic career with entrepreneurship. They recounted stories of triumphs—securing funding and commercializing breakthroughs—and difficulties, such as assembling the right team. The best technology, they agreed, is nothing without a good team to bring it to market.

Images of pathology samples from Ozcan's lens-free, holographic on-chip microscope. (Image courtesy of Aydogan Ozcan)
Developing a microbial GPS
The symposium also featured presentations on research collaborations in the Blavatnik Awards community. In a joint presentation, Rob Knight of the University of California, San Diego, and Szabolcs Márka of Columbia University presented the results of their research collaborations with Jonathan Kagan and Sarkis Mazmanian, both National Award finalists.
Márka presented behavior-tracking software he created for Drosophila flies and then adapted for mice. The software analyzes more than 50 measures of motion and can identify movement-related markers of disease and health. Manipulating the animals either physically or physiologically—adding weight to the body, impairing the sensory system, or modeling a neurological condition—produced behavioral results that could be quantified and studied.
Knight took the podium next, showing the audience a map of the bacteria that comprise the human microbiome. The distinct communities of oral, skin, vaginal, and fecal bacteria should be "about as dissimilar as a coral reef and a prairie" in a healthy human, Knight said. In a patient infected with Clostridium difficile, the disorganized microbiome and associated diarrhea can be quickly corrected with fecal transplants from healthy patients. "We need to figure out for what other diseases can we identify an unhealthy state and convert you to a healthy state by modifying your microbiome," Knight said. Imbalances in the microbiome are implicated in conditions including autism, multiple sclerosis, depression, Parkinson's disease, and posttraumatic stress disorder.
Sarkis Mazmanian, a professor of microbiology at the California Institute of Technology, is the collaborative bridge between Knight's work on the microbiome and Márka's behavior-tracking technology. Mazmanian's research explores the role of the microbiome in inflammatory illnesses. He has demonstrated the potential for microbiome manipulation to reverse autism-like symptoms in mice. "The motivation to connect microbes with behavior—with better ways of tracking both—is very compelling," Knight said.
The teams are working together to design experiments on mice raised in a germ-free environment. Combining Knight's microbiome mapping with Mazmanian's hypotheses on the role of microbes in neurological disease, the researchers plan to model various disease states and analyze the missing or dysfunctional microbiota, with an eye toward treatment. Márka's software will track animal behavior and the progress of the illnesses. "We want to figure out how to develop a kind of microbial GPS that allows us to drive around the map and figure out where we need to go," Knight said.

By combining behavior-tracking software with microbiome mapping, researchers hope to explore how missing microbes may be implicated in disease, and use the results to develop treatments. (Image courtesy of Rob Knight)
The dynamic genome of a unicellular model organism
Laura Landweber of Princeton University and Columbia University, and a 2008 regional faculty winner, discussed her research on the single-celled ciliate Oxytricha trifallax—whose complex genome and extraordinary feats of DNA assembly are a model for understanding how chromosomes break apart and reassemble in more complex organisms, including humans.
O. trifallax have multiple genomes. The macronucleus contains the working DNA used to carry out the processes of life. A germline micronucleus stores genetic information to be passed on during mating. This archival genome is heavily encrypted—gene pieces are scrambled, flipped, and rearranged in a seemingly incomprehensible jumble.
The ciliates exchange DNA during mating, and in a process that Landweber characterized as "making sense out of nonsense," each organism sorts through the jumble, discarding "sometimes as much as 90 to 95 percent of its DNA in the process of extracting the functional genetic information from the precursor genome." About 225 000 pieces of DNA are meticulously flipped, rearranged, and stitched back together to produce a new nucleus of about 20 000 single-gene chromosomes, each protected by its own set of telomeres. Millions of noncoding RNAs from the original parent nucleus drive this process, which takes just 60 hours.
Landweber explained that chromosomal rearrangement in humans is often a harbinger of cancer. O. trifallax may help scientists understand how chromosomes can be correctly reassembled.

After O. trifallax mate, a copy of the micronucleus is processed into a new macronucleus through massive cutting, reordering, and splicing of DNA. Junk segments are deleted and retained segments are placed in correct linear order. (Image courtesy of Laura Landweber)
Neural circuits and molecules in drug-seeking and depression
Mary Kay Lobo of the University of Maryland, a 2011 regional postdoctoral finalist, presented her work on the molecular mechanisms that underlie brain responses to psychostimulants—and susceptibility to depression. She reviewed the role of dopamine in the brain's reward system, explaining that drugs of abuse hijack this circuit.
The striatum region of the brain is home to medium spiny neurons (MSNs), rich in two types of dopamine receptors, D1 and D2. D1 MSNs and D2 MSNs play opposite but coordinated roles in motor function, with D1 neurons promoting movement and D2 neurons halting it. Imbalances in these receptors are associated with motor disorders such as Parkinson's disease.
Lobo was the first to examine the role of MSNs in motivation. Using optogenetic techniques, she stimulated D1 and D2 MSN activity in mice. When offered a low dose of cocaine, mice receiving D1 MSN stimulation developed drug-seeking behaviors. Conversely, even when offered high doses of cocaine, mice receiving D2 MSN stimulation showed fewer drug-seeking behaviors than control mice. Overexpression of the transcription factor early growth response 3 (Egr3) in D1 MSNs enhanced cocaine-seeking behaviors, while overexpression of the same factor in D2 MSNs suppressed the behaviors.
Lobo found similar results in depression. Stimulation of D1 activity in mice vulnerable to depression eased their symptoms, while stimulation of D2 activity in control mice resulted in avoidance in social trials. These findings indicate that the neurons that suppress drug-seeking behavior can also dampen social drive in depression.
Illustrating nanophotonics
Marin Soljačić of the Massachusetts Institute of Technology, a 2014 national laureate, delivered the final presentation of the symposium, which ended, as it began, with nanophotonics. His team has created novel nanophotonic structures with potential applications in lasers, sensors, solar cells, and microscopes. Nanophotonics has enabled highly precise control of light; scientists can "tailor the laws of physics, as far as light is concerned, almost at will," Soljačić said.
Among the advances is a method of trapping light that upends conventional techniques, which usually involve mirrors or other reflective surfaces that block light. Soljačić's approach pits wavelengths of light with opposing phases against each other such that the two cancel each other out, confining those wavelengths within the structure. He also described his work designing the first system capable of selectively filtering light by angle of propagation, rather than by color or by polarization.
Soljačić closed with a note of thanks to the Blavatnik Awards community, and announced the first annual recipient of his own philanthropic program. Modeled on the Blavatnik Awards, the prize honors an outstanding high school science student in Soljačić's native Croatia.