あなたがこの世を去って長い年月が経ったとき、あなたの科学的遺産は、主として、あなたがこの世に書き残した論文とそれらの論文が放つインパクトです。(When you are long gone, your scientific legacy is, in large part, the literature you left behind and the impact it represents.) - フィリップ・E・ボーン
The student council (http://www.iscbsc.org/) of the International Society for Computational Biology asked me to present my thoughts on getting published in the field of computational biology at the Intelligent Systems in Molecular Biology conference held in Detroit in late June of 2005. Close to 200 bright young souls (and a few not so young) crammed into a small room for what proved to be a wonderful interchange among a group of whom approximately one-half had yet to publish their first paper. The advice I gave that day I have modified and present as ten rules for getting published.
Rule 1: Read many papers, and learn from both the good and the bad work of others.
It is never too early to become a critic. Journal clubs, where you critique a paper as a group, are excellent for having this kind of dialogue. Reading at least two papers a day in detail (not just in your area of research) and thinking about their quality will also help. Being well read has another potential major benefit—it facilitates a more objective view of one’s own work. It is too easy after many late nights spent in front of a computer screen and/or laboratory bench to convince yourself that your work is the best invention since sliced bread. More than likely it is not, and your mentor is prone to falling into the same trap, hence rule 2.
Rule 2: The more objective you can be about your work, the better that work will ultimately become.
Alas, some scientists will never be objective about their own work, and will never make the best scientists—learn objectivity early, the editors and reviewers have.
Rule 3: Good editors and reviewers will be objective about your work.
The quality of the editorial board is an early indicator of the review process. Look at the masthead of the journal in which you plan to publish. Outstanding editors demand and get outstanding reviews. Put your energy into improving the quality of the manuscript before submission. Ideally, the reviews will improve your paper. But they will not get to imparting that advice if there are fundamental flaws.
Rule 4: If you do not write well in the English language, take lessons early; it will be invaluable later.
This is not just about grammar, but more importantly comprehension. The best papers are those in which complex ideas are expressed in a way that those who are less than immersed in the field can understand. Have you noticed that the most renowned scientists often give the most logical and simply stated yet stimulating lectures? This extends to their written work as well. Note that writing clearly is valuable, even if your ultimate career does not hinge on producing good scientific papers in English language journals. Submitted papers that are not clearly written in good English, unless the science is truly outstanding, are often rejected or at best slow to publish since they require extensive copyediting.
Rule 5: Learn to live with rejection.
A failure to be objective can make rejection harder to take, and you will be rejected. Scientific careers are full of rejection, even for the best scientists. The correct response to a paper being rejected or requiring major revision is to listen to the reviewers and respond in an objective, not subjective, manner. Reviews reflect how your paper is being judged—learn to live with it. If reviewers are unanimous about the poor quality of the paper, move on—in virtually all cases, they are right. If they request a major revision, do it and address every point they raise both in your cover letter and through obvious revisions to the text. Multiple rounds of revision are painful for all those concerned and slow the publishing process.
Rule 6: The ingredients of good science are obvious—novelty of research topic, comprehensive coverage of the relevant literature, good data, good analysis including strong statistical support, and a thought-provoking discussion. The ingredients of good science reporting are obvious—good organization, the appropriate use of tables and figures, the right length, writing to the intended audience—do not ignore the obvious.
Be objective about these ingredients when you review the first draft, and do not rely on your mentor. Get a candid opinion by having the paper read by colleagues without a vested interest in the work, including those not directly involved in the topic area.
Rule 7: Start writing the paper the day you have the idea of what questions to pursue.
Some would argue that this places too much emphasis on publishing, but it could also be argued that it helps define scope and facilitates hypothesis-driven science. The temptation of novice authors is to try to include everything they know in a paper. Your thesis is/was your kitchen sink. Your papers should be concise, and impart as much information as possible in the least number of words. Be familiar with the guide to authors and follow it, the editors and reviewers do. Maintain a good bibliographic database as you go, and read the papers in it.
Rule 8: Become a reviewer early in your career.
Reviewing other papers will help you write better papers. To start, work with your mentors; have them give you papers they are reviewing and do the first cut at the review (most mentors will be happy to do this). Then, go through the final review that gets sent in by your mentor, and where allowed, as is true of this journal, look at the reviews others have written. This will provide an important perspective on the quality of your reviews and, hopefully, allow you to see your own work in a more objective way. You will also come to understand the review process and the quality of reviews, which is an important ingredient in deciding where to send your paper.
Rule 9: Decide early on where to try to publish your paper.
This will define the form and level of detail and assumed novelty of the work you are doing. Many journals have a presubmission enquiry system available—use it. Even before the paper is written, get a sense of the novelty of the work, and whether a specific journal will be interested.
Rule 10: Quality is everything.
It is better to publish one paper in a quality journal than multiple papers in lesser journals. Increasingly, it is harder to hide the impact of your papers; tools like Google Scholar and the ISI Web of Science are being used by tenure committees and employers to define metrics for the quality of your work. It used to be that just the journal name was used as a metric. In the digital world, everyone knows if a paper has little impact. Try to publish in journals that have high impact factors; chances are your paper will have high impact, too, if accepted.
When you are long gone, your scientific legacy is, in large part, the literature you left behind and the impact it represents. I hope these ten simple rules can help you leave behind something future generations of scientists will admire.
There are many texts on the philosophy of science and scientific method that deal extensively with the hypothesis but, in short, we can describe it as ‘a reasonable scientific proposal’. It is not a statement of fact but a statement that takes us just beyond known information and anticipates the next logical step in a sequence of supportable precepts. The hypothesis has to have two attributes to be useful in scientific investigation: it must fit the known information and it must be testable. To comply with the first attribute, you the scientist have to read and understand the literature. To comply with the second, you have to do an experiments. In essence, the paper you are about to write concerns nothing other than those two things. You can see why the hypothesis is so central to scientific writing. (強調のため一部太字 引用元:David Lindsay, Scientific Writing = Thinking in Words, page 7)
Neuronal oscillations have been hypothesized to play an important role in cognition and its ensuing behavior, but evidence that links a specific neuronal oscillation to a discrete cognitive event is largely lacking. We measured neuronal activity in the entorhinal-hippocampal circuit while mice performed a reward-based spatial working memory task. During the memory retention period, a transient burst of high gamma synchronization preceded an animal’s correct choice in both prospective planning and retrospective mistake correction, but not an animal’s incorrect choice. Optogenetic inhibition of the circuit targeted to the choice point area resulted in a coordinated reduction in both high gamma synchrony and correct execution of a working-memory-guided behavior. These findings suggest that transient high gamma synchrony contributes to the successful execution of spatial working memory. Furthermore, our data are consistent with an association between transient high gamma synchrony and explicit awareness of the working memory content. (強調のため一部太字 引用元:Cell 157(4):845-857の要旨)
Given that LPA is known to activate cortical contractility via the Rho/Rock pathway, we hypothesized that a mechanical polarization mechanism of the cell cortex could trigger the transformation of progenitor cells into stable-bleb cells (Carvalho et al., 2013). To test this hypothesis, we treated stable-bleb cells with the Rho kinase inhibitor Y-27632 or the myosin-II inhibitor Blebbistatin (Figure 2C). Treated cells lost their characteristic polarization, supporting a critical role for Rho/Rock-mediated cortical contractility in driving stable-bleb cell transformation. (強調のため一部太字 引用元:Cell 160(4):673-685のResultsセクションの一部)
Since computational modeling suggests that a failure to segregate protein damage may result in a reduced fitness (Erjavec et al., 2008) and functions crucial for cellular fitness are often performed by parallel partly redundant pathways, we hypothesized that machineries involved in the partitioning of protein aggregates could be identified by systematically screening for genetic interactions between SIR2 and nonessential and essential genes using synthetic genetic arrays (SGA) analysis ( Tong et al., 2001 and Tong et al., 2004). Using this approach, we found that cells lacking Sir2p share many genetic interactions with the conditional actin mutant, which is a result of sir2Δ cells showing a defect in CCT-chaperonin-dependent folding of actin. (強調のため一部太字 引用元:Cell 140(2):257-267 イントロダクション)
Given the potential for nutrients to stimulate inflammatory pathways and the importance of keeping these pathways in check, we hypothesized that previously unrecognized counterregulatory mechanisms might exist to protect cells from activation of inflammatory pathways during physiological fluctuations in nutrient exposure or in nutrient-rich conditions. We reasoned that a factor participating in such a coordinating mechanism between nutrient and inflammatory responses would be expected to meet several criteria: (1) the gene product should be present in tissue types responsible for nutrient clearance from blood, such as adipose tissue; (2)expression or activity should be regulated by both nutritional and inflammatory stimuli; (3) the factor should regulate inflammatory signaling components and/or gene expression (cells or tissues lacking such a factor would then exhibit excess or prolonged inflammation in response to challenges by both nutrients and inflammatory mediators); (4) the factor should regulate cellular metabolism, and its absence should result in impaired cellular metabolic function; and (5) through regulation of metabolic function in particular cell types and organs, the factor should also impact systemic metabolism. In this study, we identify six-transmembrane protein of prostate 2 (STAMP2) as a factor meeting these criteria. (強調のため一部太字 引用元:Cell 129(3):537-548 イントロダクション)
…At present, two theories attempt to link airway epithelial ion transport to lung defense and predict how mutations in CFTR adversely affect these relationships. One theory, the “hypotonic (low salt) ASL/defensin” hypothesis, postulates that normal airway epithelia are covered by an ASL with a [NaCl] sufficiently low (≤ 50 mM NaCl) to activate defensins and create an antimicrobial “shield” on airway surfaces (31, 38 and 10). Generation of a low [NaCl], hypotonic ASL reflects selective transepithelial absorption of salt but not water from ASL, presumably a consequence of putative airway epithelial water impermeability or surface forces (31, 44 and 46). Defensin inactivation by iso- or hypertonic ASL (i.e., [NaCl] > 100 mM) in CF is postulated to reflect the Cl− impermeability of CF epithelial cells (38 and 10). The second theory, the “isotonic volume transport/mucus clearance hypothesis,” predicts that airway epithelia regulate the volume (height) of ASL by isotonic ion and water transport to optimize mucus clearance (Boucher 1994a). In CF, the rate of isotonic ion and water (volume) transport is abnormally high, which is predicted to reduce ASL volume, concentrate mucus, and lead to abnormal mucus transport and retention of tenacious mucus masses (plaques) in airways that serve as the nidus of infection (14 and 5).
… By measuring ASL ion composition, we tested key predictions of the hypotonic (low salt) ASL/defensin hypothesis: (1) that normal airway epithelia generate and maintain hypotonic (or low salt) solutions on airway surfaces, and (2) that CF airway surface liquid is isotonic (Joris et al. 1993) or hypertonic (with excess salt) (Gilljam et al. 1989). A second series of experiments tested key aspects of the isotonic hypothesis: (1) that normal airway epithelia are isotonic volume-absorbing epithelia that produce an isotonic ASL; (2) that CF airway epithelia absorb volume isotonically at abnormally high rates; and (3) that mucus transport is deranged by depletion of the ASL through excessive isotonic volume absorption. (強調のため一部太字 引用元:Cell 95(7):1005-1015のイントロダクションの一部を抜粋)
Good science would pit theory A against theories B, C, D and E with an experiment where each theory gave different predictions. (引用元:Nakagawa and Cuthill. Effect size, confidence interval and statistical significance: a practical guide for biologists. Biol. Rev. (2007), 82, pp. 591–605. doi:10.1111/j.1469-185X.2007.00027.x)
当初は仮説Aを検証すべく始めた研究だが、実際のデータを眺めてみると、仮説Bのほうがよりよく説明できるようだ――仮にこうなってしまっても、心配はい らない。すべての事項を書き出し、仮説・目的・データの最適たるコンビネーションを選びとればよい。完成した論文上での研究目的と、仕事に取り掛かる理由 付けとしての研究目的は、往々にして異なるものだ。良質な科学といえども、大抵は日和見主義的・修正主義的なものだ。(Whitesides’ Group: Writing a paper 化学者のつぶやき Chemi-station)
今回紹介する文献“Whitesides’ Group: Writing a Paper”では、世界一線級の化学者であるGeorge M. Whitesides教授が自ら、優れた論文執筆法を解説しています。具体的には『アウトライン法』――すなわち、始めに論文の枠組み(アウトライン)を作り、研究データの蓄積と同時並行して何度も修正を加えてゆく、実験データはその時々で補完、図表重視のアピールを心がけ、テキストを書くのは一番最後、という論文作成法――の提唱を通じ、研究のプロダクティビティ・時間効率の向上に効果的たる仕事術にも触れています。(http://www.chem-station.com/blog/2009/12/whitesides_group_writing_a_pa.html)
Professor Simon Peyton Jones (Computer Scientist)
How to Write a Great Research Paper
00:20 #1 Don’t wait: write
04:13 #2 Identify your key idea
07:18 #3 Tell a story
09:34 #4 Nail your contributions
16:44 #5 Related work: later
23:47 #6 Put your readers first
28:40 #7 Listen to your readers
(34:19 トークの終わり。この動画ではなぜか、再び繰り返しになっているようです。)
科研費に採択されるための最良の方法は,「書き上げた申請書を誰かに見せて添削してもらうこと」だと思う.見てもらう人が採択経験豊富な人ならなおよい.そういった人に申請書を見てもらって何度も何度も直すのが一番よい方法だ.(小噺その10:科研費に採択されるための最良の方法 Smart Lab Life 羊土社)