Interpreting Sequences from Mastodon and T. rex

COMMENTARY Social intelligence Academic greed and greatness Soot and climate 1326 1328 1333 LETTERS I BOOKS I POLICY FORUM I EDUCATION FORUM I PERSPECTIVES LETTERS A Proposal for a Decade of the Mind Initiative diseases, explain brain and mind phenomena, spur development of novel computing architectures, and enable the construction of intelligent machines. Why is a national Decade of the Mind initiative necessary now? First, rapid technological and biomedical progress of recent years make the present time ripe for breakA DEEP SCIENTIFIC UNDERSTANDING OF HOW THE MIND PERCEIVES, throughs in the study of the mind. Second, sucthinks, and acts is within our grasp. Such an understanding will have Enrich cess in this endeavor will have broad and dramatic a revolutionary impact on national interests in science, medicine, economic growth, security, and well-being. It is our belief that para- Cognitive Science impacts on the economy, national security, and our social well-being. Finally, to achieve success, a digm-shifting progress can be made now by establishing a major Neuroscience Psychology major investment in research and development will national research initiative called “The Decade of the Mind.” be required, with a time A Decade of the Mind initiative would build horizon on the order of 10 on progress of the recent Decade of the Brain Heal Cognitive Science Cognitive Science years. A Decade of the Mind (1990–99), which dramatically increased the and Medicine Computer Science Understand initiative could achieve visibility of neuroscience (1). Unlike the Protect Neuroscience Neuroscience Decade of the Brain, which focused on neurothese goals and improve science and clinical applications, the Decade of our lives and our chilDecade of the Mind. Computer Science the Mind initiative, by necessity, should be transdisciplinary and dren’s lives in ways we Engineering A multidisciplinary iniMathematics multi-agency in its approach. Success will require research that cannot now conceive. tiative to understand Neuroscience JAMES S. ALBUS,1* reaches across disparate fields such as cognitive science, medicine, the mechanisms of the GEORGE A. BEKEY,2 mind. neuroscience, psychology, mathematics, engineering, and computer Model JOHN H. HOLLAND,3 NANCY science. Additional important insights will need to come from areas G. KANWISHER,4 JEFFREY L. as diverse as systems biology, cultural anthropology, social science, 5 MORTIMER MISHKIN,6† DHARMENDRA S. KRICHMAR, robotics, and automation technology. MODHA,7 MARCUS E. RAICHLE,8 For these reasons, we believe a Decade of the Mind initiative GORDON M. SHEPHERD,9 GIULIO TONONI10 should focus on four broad, but intertwined areas (see figure). 1Senior Fellow, National Institute of Standards and Technology, Gaithersburg, MD 20899, 1) Healing and protecting the mind. Disorders of the mind affect USA. 2University of Southern California, Los Angeles, CA 90089, USA. 3University of more than 50 million Americans annually at costs exceeding $400 Michigan, Ann Arbor, MI 48109, USA, and Santa Fe Institute, Santa Fe, NM 87501, USA. billion (2). It is our obligation as scientists and health care workers to 4McGovern Institute, Massachusetts Institute of Technology, Cambridge, MA 02139, USA. 5The Neurosciences Institute, San Diego, CA 92121, USA. 6Laboratory of Neuropsychology, address these social, personal, and economic burdens on our society. MD 20892, USA. 7IBM Almaden Research 2) Understanding the mind. Although much progress has been National Institute of Mental Health, Bethesda, Center, San Jose, CA 95120–6099, USA. 8School of Medicine, Washington University, St. achieved recently in brain research, a fundamental understanding of Louis, MO 63110, USA. 9Yale University Medical School, New Haven, CT 06510, USA. how the brain gives rise to the mind is still lacking. Knowledge of 10Department of Psychiatry, University of Wisconsin, Madison, WI 53719–1179, USA. the mind’s inner workings will require new tools that can deeply *The views expressed in this article do not necessarily represent the views of the U.S. probe mental processes. Research should be encouraged on aspects National Institute of Standards and Technology, the U.S. Department of Commerce, or the of the mind believed to be uniquely human, such as the notion of U.S. government. self, rational thought processes, theory of mind, language, and †This Letter was prepared as part of the co-author’s official duties as a U.S. government higher order consciousness. employee. The views expressed in the Letter do not necessarily represent the views of the 3) Enriching the mind. A better understanding of the mind will NIMH, NIH, DHHS, or the U.S. government. enrich our lives by improving our education system at all levels, treating mental illnesses and addictions, extending the mind to new References and Notes skills, and educating the general public on legal and ethical issues 1. E. G. Jones, L. M. Mendell, Science 284, 739 (1999). 2. J. Carey, Ed., Brain Facts: A Primer on the Brain and Nervous System (Society for involving the brain and the mind. Neuroscience, Washington, DC, 2006). 4) Modeling the mind. Combining theoretical and computational 3. The authors represent the steering committee for the Decade of the Mind initiative. A methodologies with empirical findings will be crucial for healing, Decade of the Mind Symposium was held on 21 to 22 May 2007 at The Krasnow Institute understanding, and enriching the mind. Large-scale brain modeling for Advanced Study at George Mason University in Fairfax, VA, where many of the issues contained in this Letter were originally discussed. efforts will predict and diagnose disorders, test treatments for www.sciencemag.org SCIENCE VOL 317 Published by AAAS 7 SEPTEMBER 2007 Downloaded from www.sciencemag.org on April 15, 2014 edited by Etta Kavanagh 1321 LETTERS Clarifying Cougar Management in Oregon I AM CONCERNED THAT V. MORRELL’S SCIENCE Now article “Oregon cougars to be hounded” ( h t t p : / / s c i e n c e n ow. s c i e n c e m a g . o rg / cgi/content/full/2007/629/2; posted 29 June) may lead readers to misconstrue how the passage of the new law, HB 2971, affects cougar management in the state. HB 2971 does not “bring back houndhunting” or “overturn” Measure 18, which was enacted by Oregon voters in 1994. The new law simply clarifies some of the ambiguity in the original Measure 18. Measure 18 specifically allowed the Oregon Department of Fish and Wildlife (ODFW) to use “agents” to manage cougar populations, and ODFW did so in the initial years after the measure was passed. But the Department of Justice later determined that it was not clear that ODFW had legislative authority to appoint agents. HB 2971 simply makes it clear. The photo caption used with the article states that “citizens” can now use hounds when hunting cougars. Sport hunters still cannot use hounds. Only agents selected, trained, and supervised by ODFW will be authorized to hunt with hounds, and only when they are acting in official capacity to implement cougar management. CLAIR KUNKEL Acting Deputy Director, Oregon Department of Fish and Wildlife, 3406 Cherry Avenue, NW, Salem, OR 97303, USA. ERIK RIFKIN1* AND EDWARD BOUWER2 Characterizing Health Risks 1Rifkin IN THE REPORT “A COMMON VARIANT ON chromosome 9p21 affects the risk of myocardial infarction” (8 June, p. 1491; published online 3 May), A. Helgadottir et al. use relative risks to describe an association between myocardial infarction (MI) and a common sequence variant on a specific chromosome. They conclude that individuals in the population homozygous for this variant have an estimated 1.64fold greater risk of suffering MI than noncarriers and a 2.02-fold risk for early onset MI cases. Although calculating relative risks and relative risk reduction is widely used to represent experimental results, great care needs to be taken when reporting, interpreting, and characterizing health risks and benefits based primarily on relative risks. The authors appear to allude to this issue in the last paragraph of their Report by stating: “However, as the relative risks are not 1322 extremely high, it explains only a small fraction of the familial clustering of the disease and would not generate large linkage scores.” This cryptic sentence, however, falls far short of acknowledging that an understanding of absolute risks, absolute risk reduction, and the number needed to treat will be a necessary prerequisite to making informed decisions on medical intervention. There is general agreement in the scientific community that the exclusive use of relative risks distorts and often grossly exaggerates the significance of health risks and benefits (1–4). Absolute risk values provide information essential to physicians and patients alike. For example, hypothetically, an absolute risk analysis could show that 2 individuals out of 10,000 with the variant get heart disease and 1 individual without the variant gets heart disease. In this case, the absolute risk reduction would be 0.01%, 10,000 people would have to be treated to observe 1 benefit, and 99.99% of people with the variant would not benefit from intervention. The increased relative risk for individuals with the variant, in this hypothetical example, would be 100%, a much more impressive number. The public hears about health risks and benefits from many sources on a daily basis. Although many approaches may be scientifically legitimate, the misinterpretation of statistical relationships can lead to inappropriate risk management decisions. The scientific community and the public would have been better served if limitations regarding the relevance and importance of this Report’s findings were more clearly articulated. and Associates, 10 East Lee Street, #2107, Baltimore, MD 21202, USA. 2Department of Geography and Environmental Engineering, Johns Hopkins University, Baltimore, MD 21218–2686, USA. *The author is president of a private company specializing in risk assessment. References 1. L. Hembroff et al., BMC Med. Inform. Decis. Mak. 4, 20 (2004). 2. R. Gordon-Lubitz, JAMA 289, 95 (2003). 3. R. Moynihan et al., N. Engl. J. Med. 342, 1645 (2000). 4. A. Edwards et al., J. Health Commun. 6, 61 (2001). Response RIFKIN AND BOUWER FOCUS ON THE ABSTRACT of our paper, where we summarized the effect of the discovered variant by giving its estimated relative risk, something that is relatively standard when reporting results of this sort. We cannot see how our statement that “individuals in the population homozygous for this variant have an estimated 1.64-fold 7 SEPTEMBER 2007 VOL 317 SCIENCE Published by AAAS Letters to the Editor Letters (~300 words) discuss material published in Science in the previous 3 months or issues of general interest. They can be submitted through the Web (www.submit2science.org) or by regular mail (1200 New York Ave., NW, Washington, DC 20005, USA). Letters are not acknowledged upon receipt, nor are authors generally consulted before publication. Whether published in full or in part, letters are subject to editing for clarity and space. greater risk of suffering MI than noncarriers” “grossly exaggerates the significance of health risks and benefits.” Not only do we believe relative risk to be a very relevant parameter, it is also what our study design allows us to estimate directly and, hence, the appropriate value to highlight in the abstract. The Letter gives a hypothetical example where the relative risk is twofold, but the absolute increase in risk is low, presumably to highlight that here the effect is trivial. Although the latter might hold in some cases, it is far from being true in general. Consider a more concrete example. The chance of a fatal plane crash with a major commercial airline is so small that, even for a frequent flier, the lifetime chance of being involved in a fatal accident is still very small. The latter remains small even when multiplied by five, but most people will probably agree that it is a serious matter if the chance of a fatal plane crash is suddenly increased fivefold. Return to our manuscript and MI, a rather more common occurrence than a plane crash. The “cryptic sentence” Rifkin and Bouwer refer to was meant to address a more subtle issue than the difference between absolute and relative risks. Population attributable risk (PAR) of a genetic variant for a disease is defined as the fraction of cases that would be eliminated from the population if the risks of the carriers were to be reduced to that of the noncarriers. That the discovered variant has an estimated PAR of 21% for MI in general makes its impact substantial from a public health point of view. But PAR is not the appropriate measure of the contribution of a variant to the familial clustering of the disease; it is better measured by the sibling recurrence risk ratio, λs. In particular, if two variants have the same PAR, then the one that is less frequent but has a higher relative risk will have a higher λs. Also, PAR is not additive when multiple variants are considered; mathematically, there could be 10 independent variants, each with a PAR of 50%. The point we tried to make is that, although the discovered variant has a PAR of 21% for MI, it does not mean that it accounts for 21% www.sciencemag.org LETTERS of the genetic component to the disease. However, even though PAR and λs are distinct measures, they are both calculated using relative risk instead of absolute risk. modifications are possible (1) and the spectra suggest glycine as a potential hydroxylation site, we acknowledge alternative interpretations that are consistent with our spectra, the structure of collagen, and previous collagen research since α-hydroxyglycine has been reported to be unstable (2, 3). Ion trap mass spectrometers scan very fast and are highly sensitive but cannot resolve amino acids or combinations of modifications and amino acids that are near isobaric (same nominal mass), as stated in the original Report. It is sometimes difficult to determine the precise position of a modification from adjacent or nearby amino acid residues, since MS/MS spectra often lack sufficient sitespecific fragment ions (4). Hydroxylation of P to 4-hydroxyproline is a highly abundant modification that stabilizes the triple helical structure of collagen. Hydroxylation also occurs to a lesser extent on lysine (K) residues (5, 6). In type I and type II collagens, these hydroxylation sites have been reported to exist nearly exclusively for P or K in the Y position of the collagen triplet repeat –GXY- (7, 8). A singular exception, one P in human collagen I and II, is X position hydrox- KARI STEFANSSON AND AUGUSTINE KONG deCODE genetics, Sturlugata 8, IS-101 Reykjavik, Iceland. Interpreting Sequences from Mastodon and T. rex J. ASARA ET AL. REPORTED THAT COLLAGEN proteins from well-preserved ancient fossil bones from a 160,000- to 600,000-year-old mastodon and a 68-million-year-old T. rex can be extracted and sequenced (“Protein sequences from mastodon and Tyrannosaurus rex revealed by mass spectrometry,” 13 April, p. 280). Tandem mass spectrometry (MS/MS) is an effective sequencing method for ancient fossils when DNA is not available. It has come to the original authors’ attention that there are concerns regarding the reported sequences containing glycine (G) hydroxylation, as well as some positions of proline (P) hydroxylation. Although nonstandard postmortem ylated to 3-hydroxyproline (9). For several spectra, we determined that G was hydroxylated when an alternative interpretation based on the same spectra could accommodate the more likely Y position. In another situation, we could not differentiate between isoleucine (I)/leucine (L) and hydroxylated proline, P(OH) (nominal mass of 113 Da with an exact mass difference of 0.0364 Da). Serine (S) residues adjacent to or near unmodified proline -SP- could be interpreted as alanine (A) in place of serine adjacent to hydroxyproline -AP(OH)- with the same nominal mass of 184 Da. Alternatively, –G(OH)A- could be interpreted as –GS-. Deamidation of asparagine (N) or glutamine (Q) is a posttranslational modification resulting in an amino acid mass increase of 1 Da to aspartic acid (D) and glutamic acid (E), respectively, and could not always be distinguished due to the ion trap’s resolution and mass accuracy. We have determined that one of the reported T. rex spectra for the peptide GLVGAPGLRGLPGK is statistically insignificant when searched against large protein databases and is a low confidence sequence, while the other six T. rex sequences remain high confidence. Since all sequences ✐ Searching for some fresh ideas about science education? Η 6 Η 5 Τ . − . 4 Τ Ι 5 ∀ Λ Υ Ξ ∀ Λ 6 Η ? Α 6 Λ ∀ + Ν − . 6 ς < ϑ + − 5 Α 5 Σ ∀ + ; ? ∗ Η Σ ϑ ∀ Λ Λ 6 Ξ Λ 4 . Ν Ν 5 ; + Α Λ Α Ν ? ? . . Α Α ϑ − = = Υ < Κ Τ Λ ? . > Κ Κ . Κ Κ ∀ Λ Ι Υ Τ + 5 ∀ . ϑ Λ . − Η Κ 6 ϑ ς 5 5 > 6 ? ϑ + Κ Σ > Κ 5 ϑ Λ 5 ς ∀ Σ Ξ . 6 Α 5 5 ϑ Α > Υ ? ς ; Η Ξ ς Λ ? ϑ ∗ Α 6 Α + ; . Σ Ι ϑ ; Ν 5 5 + Α ∀ Τ Η Ι Λ − ; = . < Κ = Λ > . Λ = 4 Κ < Ν 5 Ι Ν ? < ∗ 6 4 Λ Ν ? Λ Σ 4 Λ Α − . ? 4 > 4 Ν Α Υ . . Α ∀ Κ + Find answers in Science’s Education Forum. 4 < ς Η 5 Ν Τ Ι ∗ Λ Ξ > ? ? Υ The Science Education Forum is a dynamic source of information and new ideas on every aspect of science education, as well as the science and policy of education. The forum is published in the last issue of every month and online, in collaboration with the Howard Hughes Medical Institute. Keep up-to-date with the latest developments at: www.sciencemag.org/education What’s your perspective? Do you have ideas or research you’d like to share in the Science Education Forum? We’re now looking for thoughtful, concise submissions (around 2,000 words) for 2007. To submit your paper, go to: www.submit2science.org Hidden Words: assessment; education; graduate; mentoring; science; classroom; math; laboratory; outcomes; student; diversity; faculty; patnerships; tests; engineering 1324 7 SEPTEMBER 2007 VOL 317 SCIENCE Published by AAAS www.sciencemag.org LETTERS matched to a single bone protein family when searched against the all-taxon NCBI nonredundant protein database containing more than five million entries, rates of false positive identification were very low. Surprisingly, a likely alternative interpretation leads to the unique T. rex peptide sequence GAPGPQGPSGAP(OH)GPK. Alternate interpretations for the collagen sequences reported in the original manuscript are available in the Supporting Online Material (10), the sequences for mastodon and ostrich (not originally published) appear in UniProt Knowledgebase, and the complete list is being placed in a supplementary repository (ftp:// ftp.ebi.ac.uk/pub/databases/supplementary). These alternate sequences strengthen our assertion that collagen has been sequenced from ancient fossil bones without contradicting wellestablished structures for collagen modifications. In retrospect, prior knowledge of collagen structures would have helped to construct our peptide library; however, the oversight inadvertently provides us assurances that genuine collagen sequences were detected. Overall, these possible minor sequence alterations do not alter our original conclusions that ancient collagen peptides were sequenced from well-preserved mastodon and T. rex fossil bones, and that T. rex sequences match better to chicken than any other single organism of currently known sequence. For future sequencing efforts, ultra–high-resolution Fourier transform or Orbitrap mass spectrometry technology will be used for acquiring precise masses and distinguishing near isobaric amino acids (11). JOHN M. ASARA,1,2 JOHN S. GARAVELLI,3 DAVID A. SLATTER,4 MARY H. SCHWEITZER,5 LISA M. FREIMARK,1 MATTHEW PHILLIPS,1 LEWIS C. CANTLEY1,6 1Division of Signal Transduction, Beth Israel Deaconess Medical Center, Boston, MA 02115, USA. 2Department of Pathology, Harvard Medical School, Boston, MA 02115, USA. 3EMBL Outstation, European Bioinformatics Institute, Cambridge CB10 1SD, UK. 4Department of Biochemistry, Cambridge University, Cambridge CB2 1QW, UK. 5Department of Marine, Earth and Atmospheric Sciences, North Carolina State University, Raleigh, NC 27695, USA. 6Department of Systems Biology, Harvard Medical School, Boston, MA 02115, USA. References and Notes 1. 2. 3. 4. N. Tuross, Archaeometry 44, 427 (2002). A. J. Hoefnagel et al., J. Org. Chem. 57, 3916 (1992). J. H. Highberger et al., Biochemistry 21, 2048 (1982). D. T. McLachlin, B. T. Chait, Curr. Opin. Chem. Biol. 5, 591 (2001). 5. J. P. Orgel et al., Proc. Natl. Acad. Sci. U.S.A. 103, 9001 (2006). www.sciencemag.org SCIENCE VOL 317 Published by AAAS 6. A. Ayad et al., The Extracellular Matrix Facts Book (Academic Press, San Diego, CA, ed. 2, 1998). 7. T. Pihlajaniemi, R. Myllyla, K. I. Kivirikko, J. Hepatol. 13 (suppl. 3), S2 (1991). 8. M. Pekkala et al., J. Biol. Chem. 279, 52261 (2004). 9. K. Tryggvason, J. Risteli, K. I. Kivirikko, Biochem. Biophys. Res. Commun. 76, 275 (1976). 10. The Supporting Online Material is available at www. sciencemag.org/cgi/content/full/317/5843/1324/DC1. 11. J. V. Olsen et al., Mol. Cell. Proteom. 4, 2010 (2005). 12. J.M.A. acknowledges B. Brodsky for helpful discussions. J.S.G. and D.A.S. acknowledge N. Kelleher and K. Lilley for helpful discussions and P. Browne for assistance in preparing the UniProt entries. CORRECTIONS AND CLARIFICATIONS Reports: “The Release 5.1 annotation of Drosophila melanogaster heterochromatin” by C. D. Smith et al. (15 June, p. 1586). The affiliation for ShengQiang Shu and Christopher J. Mungall was listed incorrectly. They are affiliated with Berkeley Bioinformatics and Ontologies Project, Lawrence Berkeley National Laboratory, Berkeley, CA 94720, USA. In addition, two funding sources were omitted from the acknowledgements note. The work in this paper was also supported by NIH grant P41 HG000739-15 (S.S.) and by the Howard Hughes Medical Institute (C.J.M.). Special Issue on Sustainability and Energy: Perspectives: “Ethanol for a sustainable energy future” by J. Goldemberg (9 February, p. 808). There are numerical errors in Table 1. Under the heading “Geothermal energy,” the value for “Total” should be 1.08, and the value for “Heat” should be 0.80. 7 SEPTEMBER 2007 1325