This article is the backstory behind: S. Fritz, L. See, I. McCallum et al., Mapping global cropland and field size. Global Change Biology, 21(5), 1980-1992. doi: 10.1111/gcb.12838  

This is the backstory behind the scientific article entitled "Climate-related range shifts – a global multidimensional synthesis and new research directions" published in Ecography \citep*{Lenoir2015}.

This article is the backstory behind: S. Van den Bosch, W. Schutyser, B.F. Sels, et al., Reductive lignocellulose fractionation into soluble lignin-derived phenolic monomers and dimers and processable carbohydrate pulps \cite{Bosch2015}. 

The critical event that eventually led to the first of my meta-analysis papers on genetic rescue occurred in February 2007 at a book writing session on the second edition of “Introduction to Conservation Genetics” \citep{Frankham} at Jonathan Ballou’s house in the Washington, D.C. area. Upon reaching the topic of outbreeding depression (where the effects of crossing populations results in harmful fitness effects in the progeny), we both expressed serious disquiet that the risks of outbreeding depression were being overplayed, while the potential fitness benefits of crossing (genetic rescue) were largely being ignored. One of us said “we must be able to predict the risk of outbreeding depression”. A few days later inspiration struck and we had the key to doing this: harmful effects on fitness of crossing populations typically arise when the crossed populations have fixed chromosomal differences, and/or are adapted to different environments. We subsequently recruited Katherine, Ralls, Mark Eldridge, Michele Dudash, Charles Fenster and Robert Lacy and jointly transformed this insight into a paper that was published in Conservation Biology \cite{FRANKHAM_2011}. That work was critical to the ability to use genetic rescue (variously called outcrossing or augmentation of gene flow) as a tool to save small inbred population fragments from extinction, and thereby reduce population and species extinction risks. As genetic rescue had been attempted in very few cases, we decided to write a book on “Genetic Management of Fragmented Animals and Plant Populations” in an attempt to create a paradigm shift where the discovery of genetically differentiated populations was followed, not by the conclusion that separate management of fragments was required, but by asking if any of the populations were suffering genetic erosion (inbreeding, loss of genetic variation, reduced fitness, reduced ability to evolve and elevated extinction risk), and if so, was a genetic rescue attempt justified. I drafted Chapter 6 on Genetic rescue for the book, and then decided that it needed some examples which were put into a Table. At this point, I finally recognized that a fully-fledged meta-analysis was required, as there was no overview of the effects of outcrossing in a conservation context, i.e. when an inbred population fragment with low genetic diversity was crossed to another population and where the risk of outbreeding depression in the resulting progeny was low. The meta-analysis was done without external research funding as I have been officially retired since 2002 (but am still scientifically active) and do not have grant money for any of the work described here. I am great fan of meta-analyses: not only can they be done without research funds, but they are typically highly cited, similar to reviews, and are superior scientifically to them. By mining the literature, I found 156 relevant comparisons of inbred parents and their outcrossed progeny, and 145 had beneficial effects on fitness. Only one of the cases where crossing was harmful was a convincing case of outbreeding depression (in a selfing nematode), the others likely being chance observations due to low statistical power. The median fitness benefit from augmenting gene flow was 148% in wild/stressful conditions and 45% in benign/captive ones. Consequently, there are huge potential benefits from augmenting gene flow into population fragments suffering from genetic erosion, provided the risk of outbreeding depression in proposed crosses is low. Thus, the two main impediments to genetic rescue attempts have been removed. This paper was published in Molecular Ecology \cite{Frankham_2015} (currently 123 citations in Google Scholar), and was accompanied by a commentary from Donald Waller \cite{Waller_2015}. He praised the paper, but was not convinced about the persistence of the benefits over generations. Consequently, I did further analyses on my database to compare the effects of crossing on fitness in the F1, F2 and F3 generations and this confirmed that the benefits persisted to an extent that was, if anything, better than expected. This led to the publication of a second genetic rescue meta-analysis paper in Biological Conservations \cite{Frankham_2016}. Writing of our book continued (with Paul Sunnucks being added as another author) and it was submitted to Oxford University Press in December 2016. However, during the subsequent copy editing I realised that the second genetic rescue meta-analysis paper was incomplete, as the persistence of fitness benefits following crossing is expected to depend on the breeding system. Persistence of fitness benefits across generations is expected for outbreeders, but habitual selfing after crossing will lead to loss of benefits, while mixed mating species should experience only partial persistence of fitness benefits. I subsequently extended the analyses of my database from F3 to F13 and found no significant decline in fitness benefits for outbreeding species. Further, \citet{Bijlsma_2010} found no significant change in fitness between F10 and F15 generations in outbreeding Drosophila flies. The updated findings were included in the published version of our “Genetic Management of Fragmented Animal and Plant Populations” book \citep{Frankham_2017}. This was followed by a related paper calling for a paradigm shift in the genetic management of fragmented populations \citep{Ralls_2017}.

This is the Backstory behind Current developments in fluorescent PET (photoinduced electron transfer) sensors and switches \cite{Daly_2015}

This is the backstory of: From industrially weavable and knittable highly conductive yarns to large wearable energy storage textiles \cite{Huang_2015b}.

This is the Backstory of  Nanopatterned Textile-Based Wearable Triboelectric Nanogenerator \cite{Seung_2015}.

This is the backstory of:  A randomized, double-blind phase 2 clinical trial of blosozumab, a sclerostin antibody, in postmenopausal women with low bone mineral density \cite{Recker_2015}Lilly conducted a clinical trial of Blosozumab, and before it was approved by the FDA, they abandoned it completely.

Backstory of  Highly efficient Cas9-mediated gene drive for population modification of the malaria vector mosquito Anopheles stephens \cite{Gantz2015}. Developing gene-drive technology:  In late 2014 Valentino Gantz, then a graduate student in my lab at the University of California, San Diego (UCSD), described to me an idea for a CRISPR-based strategy to uncover the effect of recessive mutations in a single generation by creating a self-propagating mutagenic genetic element.  In our discussion, it became clear that this element might also propagate itself through mating and be passed on to the next generation at frequencies much higher than predicted by  standard "Mendelian" inheritance, a phenomenon often referred to as gene-drive1.  Following a series of additional discussions within the lab and advisory faculty members at UCSD, including Marty Yanofsky (then chair of CDB), Bill McGinnis (Dean of Biological Sciences), Steve Wasserman, and Joseph Vinetz (member of UCSD Institutional Biosafety Committee), we planned, constructed, and safely tested (in Joe Vinetz's ACL2 insectary) this genetic system in the fruit fly Drosophila melanogaster.  We called this method the mutagenic chain reaction (MCR) in analogy to the polymerase chain reaction (PCR).  We now refer broadly to the use of self-propagating CRISPR-based genetic elements as "active genetics". We submitted a manuscript for publication in Science on December 31, 2014, describing a simple proof-of-principle experiment in which an MCR element inserted into the Drosophila yellow gene, which is required for body pigmentation, was transmitted to nearly all offspring.  This paper, which was published online on March 19, 2015 \cite{Gantz_2015}, attracted considerable media attention and is credited with being the first to experimentally demonstrate a CRISPR-based gene-drive in a multicellular organism with a dedicated germline.  The study was cited as one of the primary reasons for CRISPR being chosen as the breakthrough of the year in 2015 by Science magazine.  This discovery also was ranked #11 overall and #2 in biology in the 100 top discoveries of 2015 by Discover Magazine.Applying a gene-drive system to combat malaria:  One application of gene-drive technology under discussion since the mid-1960s \cite{CURTIS_1968} is to use one of many possible strategies  to generate an inheritance bias and spread a desired trait into a target population. For example, a gene-drive element could be used to either kill mosquitoes as a kind of genetic insecticide \cite{Burt_2003}, or to spread a trait throughout a mosquito population that rendered them incapable of transmitting malarial parasites \cite{2006}.   Anthony (Tony) James at the University of California at Irvine (UCI) is one researcher who has dedicated his career to combatting malaria.  Tony was the first person ever to clone and describe a protein-encoding gene from any mosquito \cite{James_1989} and in his lab he first developed reliable methods for producing transgenic mosquitoes \cite{Jasinskiene_1998}.  His group then engineered sophisticated genetic cassettes that produce single chain antibodies that were capable of fully blocking transmission of malarial parasites \cite{Isaacs_2012,Isaacs_2011}. These antibodies bind to and neutralize malarial parasites following a female mosquito feeding on blood (only female mosquitoes bite).  James proposed to use a gene-drive approach to spread the immunizing antibody cassettes into mosquito populations, which should render those mosquitoes incapable of transmitting malaria to humans.  The final prescient sentence of his most recent paper on that topic published in 2012 stated: "If coupled with a mechanism for gene spread, antibody-expressing, malaria-resistance transgenes could become a self-sustaining disease control tool" \cite{Isaacs2012}.  On January 9, 2015, shortly after submission of our paper to Science describing efficient transmission of the MCR element in Drosophila, Valentino and I contacted Tony to ask him if he might be interested in a collaboration in which we hooked up our gene-drive system to his "immunizing" gene cassette.  Tony enthusiastically agreed.  Based on subsequent discussions, Valentino designed and constructed an immunizing gene-drive element targeting insertion into a gene called kynurenine hydroxylase (kh), which is required for normal eye pigmentation.  Valentino sent the completed kh-MCR construct to the James group, who then injected this synthetic DNA into mosquitoes.  On July 29, 2015 the James team recovered 2 transgenic individuals carrying the kh-MCR construct (out of ~ 25,000 injected larvae screened).  The memorable email that Tony sent me reporting the first results indicating how the kh-MCR was transmitted to progeny was entitled: "Friar Mendel takes a break".  After analyzing the transmission of the kh-MCR element over four generations, we prepared a manuscript entitled "Highly efficient Cas9-mediated gene drive for population modification of the malaria vector mosquito Anopheles stephensi" \cite{Gantz08122015} , which is the subject of this Backstory.  This paper, published on November 23, 2015 in the Proceedings of the National Academies of Science (PNAS), described the highly efficient transmission of the kh-MCR element to >99% of offspring via male parents.  There was significant media coverage of this first demonstration of efficient CRISPR-based gene-drive in mosquitoes, which is still ongoing over two years later. Addressing safety, ethical, and societal aspects gene-drives:  Given the potential of gene-drive systems to alter wild populations, Valentino, Tony, and I have devoted significant effort to engage scientists, social scientists, governmental regulators, and the public to initiate discussion regarding the ethical use of gene-drive systems to benefit society.  These efforts have encompassed areas including ethics, political science, sociology, health policy, as well as government regulatory agencies to provide transparent and informative explanations of active genetics technology and it applications.  For example, we jointly co-authored an opinion piece in Science with 23 other colleagues suggesting precautionary measures for safe use of gene-drive systems in the laboratory \cite{Akbari_2016} and organized a UCSD/JCVI workshop on regulation of gene-drives \cite{Adelman_2017}.  We also provided briefs to the NAS and federal risk assessment agencies (e.g., an Obama administration risk assessment panel comprising members of the FBI, CIA and Department of Homeland Security as well as presentations to the JASON study group) and a series of presentations to university Biosafety (e.g., IBC) and bioethics groups.  We continue to participate in forums for scientific and community engagement and contribute actively to ongoing discussions on these important topics. A future vision for active genetics:  In January 2016, Valentino Gantz and I published our vision for the future of active genetics in a BioEssays review \cite{BIES:BIES201500102}.  In this review, we outlined applications of active genetics to gene-drive systems, auxiliary updating drive elements such as CHACRs, reversal drives such as ERACRs and eCHACRs, bi-partite trans-complementing drives, split-drive CopyCat site-directed transgenesis vectors, potential uses for human cell therapy applications, and a summary of the ethical and safety considerations associated with gene-drive technology.Acknowledgements The author thanks Anthony (Tony) James) and Valentino Gantz for helpful comments on this Backstory and for the joy of our shared adventure 1Early geneticists referred to this phenomenon as ‘meiotic drive’ 

HistoryI decided to be a scientist when I was 7 years old after watching Carl Sagan on TV talk about neutron stars. By 16 I discovered the protein folding problem: a disarmingly simply stated problem whose solution lay at the heart of everything in biology. How hard could it be?So a degree in Biochemistry later, I did a PhD in the lab of a computational molecular modeller (Mike Sutcliffe) and got started on the problem of clustering protein structures from NMR. I wrote my first program in C and built my first web site with my characteristically stupid pun in the name - OLDERADO: On-Line Database of Ensemble Representatives and Domains [1] - to look at the results. Each morning I would go on the internet as it then was to check the dozen or so new sites that had sprung up the night before and which had been manually added to a list known as the ‘internet directory’ by someone ‘out there’ on the new web. Google appeared about a month later. Websites were going to be a big deal. Being an early adopter of web-based science seemed like a good bet.Then a move to London to be with Mike Sternberg, a world expert on protein structure prediction. I inherited an early protein fold recognition program (Foldfit [2]) written by Rob Russell, Mansoor Saqi, Paul Bates and Roger Sayle (of Rasmol fame) and had to improve it. The result was 3D-PSSM [3] and I made a ‘fancy’ website (read: 1990’s black background with gold, animated logo) for other people to use the program and for me to figure out where the bugs were. It worked great - we won the automated section of the international CASP competition in structure prediction, and hundreds of people started using the site. It was one of the first of its kind, which made it easier to stand out. Throughout I was mainly making the site to help me understand how my software worked in various situations. With so many different types of data to look at, I needed a way to sensibly display this info just to understand what was going on internally. This remains a critical focus.New techniques arose in the field and I sat on my laurels with 3D-PSSM. I neglected the site and realised only later how valuable a large internet audience is and how I had squandered it at its peak. So I vowed to try again and rebuild it from the ground up, this time properly. Using newfound skills in web development I created Phyre [4]. This did well in both scientific recognition and number of users so I was very pleased. But soon funding looked iffy and was about to run out. Funding I wondered if any users could help. So I asked on the Phyre page if anyone could write a letter of support for us that we could include with our application for funding. We received over 1,000 letters of support which was amazing and had a huge effect on our application. But I think it placed the funding bodies in a difficult position. Making web sites that help scientists use state-of-the-art tools is undoubtedly a good thing. These site are continuously working, 24 hrs/day reliably doing some work to help science. They seem like safe and sensible places to direct funding - IF they are widely used. IF they are widely used, the funding body would look deliberately negligent if they failed to fund it. But at the same time they are justified in thinking this is outside their remit. They decide on science funding, not tools. So new arms of the funding bodies started to form, dedicated to tools and resources. This is where we are today. But deciding how to handle funding for tools is a new challenge we are all still grappling with. The quest for funding is ever-present. It is critical that tools are maintained. How do you justify the maintenance cost? It is easy to fall through the funding cracks - it’s neither proper science, nor pure infrastructure. I get very paranoid about funding living in this limbo. But if user numbers and citations continue to increase, we are safe. These are the only ways to prove your value in this game. But this hopefully reflects real-world utility of the tools you create. So what’s the best way to maximise the utility of these tools to researchers?Build it, and they will comeGreat tools based on state-of-the-art algorithms often go unused and unnoticed - great science gets ‘hidden’ in the hands of a few people competing intensely over small absolute changes in accuracy using programs designed for no one but the developers. The gains to society and science by enabling more people to use these tools are far greater. But then they need to be far easier to use if more people are to use them. Hence my focus on user interfaces. The best way to learn what works for user interfaces is by using lots of them. And thinking about the types of question a user may have. Simply BE a user and every time you want to do something and CAN’T, write a program to do it. Accumulate these programs in a web site. This empowers me and anyone else who cares, with these new tools - a virtuous circle.‘BE the user’ sounds great. But it gets harder as you spend more time with computer development and less time working on a biological problem with real proteins. I don’t mean wet lab. Just computational analysis problems with specific proteins faced by real scientists. That’s the hardest part to remain connected to. For now I just imagine scenarios. I imagine all the combinations we can make by connecting our tools in different ways, and build some that seem most promising. I meet users at workshops to find out what they want. But this is where the next development needs to happen. How can I communicate effectively with 50,000 users about what they want and what I can deliver? This is the biggest challenge for me right now.Despite the funding complexities, we got funding for Phyre2 [5]. THIS time I can do it properly, I thought. Use the most up to date facilities in the browser to make it look nicer, easier to use, and add those new ideas I had or were suggested to me by users at workshops (PhyreAlarm, BackPhyre, One-2-one threading). I also tried to improve how a user could look both at sequence and structure at the same time in the browser to analyse a range of features. This led to the development of Phyre Investigator (one module in Phyre2) and honed my skills at javascript for the future. This most recent paper has again done well in the citation game which I hope reflects its usefulness to researchers ‘out there’ on the web.FoldingIt’s important to remember the reason I’ve done any of this, and that reason is the mystery of protein folding. The person that solves the protein folding problem (at least I hope it’s a person) won’t do it with pen and paper. It will be someone extremely adept at using a variety of tools that probably already exist individually, and who will put them together in the right way. That is generic problem solving. Make this easier for people, and problems in general get solved faster. With folding my main fear is this being one of the first major science problems where artificial intelligence gets there before us. Google DeepMind is after it, and we’ve seen what they’ve done to the game Go.Protein folding is the molecular biology equivalent of the Goldbach conjecture: The impossible task that no one would sensibly pay you to work on but you want to work on because of its elegant simplicity and deep importance. So you need to do something useful while you ponder the impossible. You find the most useful thing you can do for everyone else and for you. Keep it as close as you can to folding whilst maintaining an audience size that justifies your existence - i.e. homology modelling. Problems that no one has been able to solve for decades typically don’t fall over from expected directions of attack. To be a long-standing problem means most lines of attack that occur to people have been tried. So new lines must be found by looking further afield, in other areas of science where an analogous problem may have been thwarted. But to see the mapping between a problem in say physics or economics or mathematics, to the folding problem requires a decent understanding of both areas. So that means trying to learn physics, signal processing, quantum mechanics, maths, computer science, AI, etc. All in the hope of seeing a new angle from which to attack the problem. And it’s fun and fascinating. But you need to make yourself useful to the world and justify your salary. My most recent direct stab at folding involved eigen decomposition of protein contact maps. As you may surmise, I have not yet been contacted by Stockholm.Where to publish?Well it’s not really science is it? Its tools to do science. So it doesn’t sit easily with most journals. You can submit an ‘Application Note’ which is about a page describing your web server. This doesn’t really count as a proper scientific paper in many eyes. Or often authors will use their tool to do some rapidly thrown together analysis that makes what would otherwise be an application note pass for a normal scientific article. Neither of these scenarios is ideal. The way I see it, when I’ve made a tool I haven’t answered a biological question, but I’ve made it easier to answer many biological questions. In the end we went for Nature Protocols which is aimed at step by step instruction for how to use previously published tools. It fits, albeit somewhat uncomfortably.FutureSo Phyre2 has done very well in attracting users and now I’m on to Phyre3 which will be out shortly (end of 2017). Again it’s a full redesign, this time with two people on it, me on the front end and my colleague Stefans Mezulis on the back-end, using the browser to its best, new web tech, polished with a new engine (the PhyreEngine, naturally) and more hardware for faster processing of bigger genomes. Also an entirely new tool called PhyreRisk is being developed: a portal for analysing disease, mutations, structures and complexes. But the primary focus is still the same: utility. What do people want or need to do and what can be done computationally to help them. The better we can match what can be done computationally with what researchers want to do, the faster we will make progress in all of science. And maybe we’ll beat DeepMind to the answer to protein folding. ReferencesOLDERADO: On‐line database of ensemble representatives and domains. LA Kelley, MJ Sutcliffe, Protein science 6 (12), 2628-2630 (1997)Recognition of analogous and homologous protein folds - assessment of prediction success and associated alignment accuracy using empirical substitution matrices. Russell RB, Saqi MA, Bates PA, Sayle RA, Sternberg MJ. Protein Eng. Jan;11(1):1-9. (1998)Enhanced genome annotation using structural profiles in the program 3D-PSSM. LA Kelley, RM MacCallum, MJE Sternberg. Journal of molecular biology 299 (2), 501-522. (2000)Protein structure prediction on the Web: a case study using the Phyre server. LA Kelley, MJE Sternberg. Nature protocols 4 (3), 363-371. (2009)The Phyre2 web portal for protein modeling, prediction and analysis. LA Kelley, S Mezulis, CM Yates, MN Wass, MJE Sternberg. Nature protocols 10 (6), 845-858 (2015)

This article was originally published at The Winnower and is the backstory behind Cook, J., D. Nuccitelli, S.A. Green, M. Richardson, B. Winkler, R. Painting, R. Way, P. Jacobs, and A. Skuce. 2013. "Quantifying the consensus on anthropogenic global warming in the scientific literature." Environmental Research Letters no. 8(2) (024024). doi: 10.1088/1748-9326/8/2/024024.In 2004, Naomi Oreskes published a paper in Science investigating the level of agreement in the peer-reviewed literature on climate change (Oreskes 2004). Among the 928 'global climate change' papers published between 1993 to 2003, not a single paper rejected the scientific consensus that humans were causing global warming. This result was featured by Al Gore in his Oscar winning film, An Inconvenient Truth.This was the first of several papers investigating scientific consensus among climate scientists. A survey of over 3000 Earth scientists found that the higher the expertise in climate science, the greater the level of agreement on humanity's role in climate change (Doran and Zimmerman 2009). Among climate scientists actively publishing climate research, 97% agreed that humans were significantly changing global temperature.A similar result was obtained two years later in a study collating public declarations on climate change (Anderegg et al. 2010). Among scientists who had published peer-reviewed research on global warming, around 97% endorsed the consensus.When you consider that not a single national scientific organization in the world disputes the consensus, the conclusion is inescapable. The scientific agreement that humans are causing global warming is overwhelming and robust.However, this conclusion has yet to seep into public consciousness. A 2012 survey of Americans found that 57% of the public disagree or are unaware of the scientific consensus (Pew 2012). There is a gaping chasm between public perception of consensus and the 97% reality.In light of this, we decided to update Oreskes' examination of the peer-reviewed literature by adding another decade of peer-reviewed research to the analysis. The period examined was 1991 to 2011. We also broadened the search by adding 'global warming' papers (Oreskes originally searched only 'global climate change' papers). The larger sample of over 12,000 papers allowed us to examine the evolution of consensus over the 21-year period.Research was done by a team of volunteer researchers, all of them members of the Skeptical Science team. Skeptical Science is a website whose purpose is to refute climate misinformation with peer-reviewed science. This was a citizen science effort with a mix of team members, including a diversity of academics from all over the world. This included a climate scientist from Canada, chemistry and environmental scientists from the USA, a meteorologist from England and yours truly, a social scientist from Australia.The initial challenge was maintaining motivation among the volunteers over several months, as we read and rated over 12,000 abstracts (not just once but twice in order to double check each others' ratings). The work was tedious but not without its rewards - one rater reported feeling giddy when he encountered an abstract rejecting the consensus! I later learnt that there had been two similar efforts to measure the level of consensus in published climate papers. In both cases, the sample size was around 1000 papers. Both efforts were never completed. Had I known this, perhaps I wouldn't have been so ambitious with our sample size.Another hiccup during the process was our website getting hacked (Lacatena 2014). Rather than cope with multiple email threads, we collaborate and discuss our research in a private web-based forum. Midway through the rating process, the forum was hacked and published online. This led to a number of bloggers poring over years' worth of private conversations, including discussions about our consensus research. The research was suspended for a short period while we relocated our forum and the online abstract rating system.Our paper was rejected by the first two journals we submitted to. The reason? Establishing scientific consensus is stating the obvious. To paraphrase the reviews: tell us something we don't know! We worked hard to overcome this potential difficulty, including novel methods and results in our analysis. We invited authors of the climate papers in our analysis to categorize their own research, providing an independent measure of consensus. In our resubmission, we were careful to more strongly emphasize the novel aspects of our research.Our paper was accepted by Environmental Research Letters, a journal launched in 2006 that already has an impact factor rivaling such prestigious journals as Geophysical Research Letters and Climatic Change. One of our priorities was making our paper open-access. Hiding it behind a pay wall would have been hideously unproductive, given our aim of increasing public awareness of consensus. The fee to make the paper open-access was USD$1,600. We put the call out for PayPal donations in a Skeptical Science blog post (Cook 2013a). It took 9 hours to raise the funds.Once our paper (Cook et al. 2013) was published, our results generated a great deal of interest. Distributing press releases from the universities of myself and my coauthors led to media coverage in over 28 countries, including a number of non-English speaking countries (thank goodness for Google Translate). Interest peaked when President Obama's Twitter account tweeted about our research to 31 million followers (Obama 2013b). Several weeks later, President Obama mentioned the 97% consensus in a landmark speech on climate change (Obama 2013a).Altmetrics is a website that measures chatter about scholar articles by monitoring mainstream media and social media outlets such as blogs and Twitter. Our paper was ranked as the 11th most talked about scholarly paper in 2013 (Altmetric 2014). Sadly, we were pipped out of the top ten by a paper about Sudoku. Our research was also listed in the top 5% of all scholarly papers. This last statistic, ranking the level of social media chatter on all academic papers ever published, does need to be taken with a grain of salt. There was no one in 1859 tweeting "OMG! Tyndall just discovered GHGs!"Not all of the chatter was glowing. In less than a year, over 200 attacks on our research have been published. There have been blog posts, cartoons, reports, a YouTube video, a paper, TV interviews, and testimonies in government hearings. The one thing these attacks have in common is almost all were conducted by individuals or groups who reject the scientific consensus on human-caused global warming.Why did our research provoke such strong reaction? Ours is the latest episode in over two decades of attacks on scientific consensus. In 1991, a fossil fuel group spent $510,000 on a PR campaign to reposition global warming as theory, not fact (Oreskes 2010). In recent years, the most common climate myth promoted by conservative columnists was "there is no consensus" (Elsasser and Dunlap 2013). Attacking the consensus is a central strategy by opponents of climate action.Our research has also withstood some criticism from several academics that don't dispute the consensus. In these cases, I've been diligent in promptly and publicly responding to comments on our research by respected academics.Law and psychology professor Dan Kahan questioned whether communicating the scientific consensus was an optimal approach, given that it may provoke a negative reaction from those dismissive of climate change (Kahan 2013). In response, I outlined the evidence for both the efficacy and importance of consensus messaging (Cook 2013b). Consensus information increases both acceptance of human-caused global warming (Lewandowsky, Gignac, and Vaughan 2013) and support for climate policy (Bolsen, Leeper, and Shapiro 2013). But crucially, consensus messaging was shown in an Australian experiment to partially neutralize the biasing influence of ideology with conservatives showing a greater increase in belief in human-caused global warming than liberals (Lewandowsky, Gignac, and Vaughan 2013). A study with U.S. participants found that the increase in perceived consensus in response to consensus information was greatest among conservatives (Kotcher et al. 2014).Climate scientists at the Climate Change National Forum also questioned the value of the research, arguing that scientists should be discussing more nuanced questions than human-caused global warming (Forum 2014). In response, I pointed out that while the fundamental fact that humans are causing global warming was considered settled by scientists decades ago, the general public continue to be confused over even these fundamental elements of climate science (Pew 2012).Climate scientist Mike Hulme argued that instead of discussing consensus, we should be talking about policy options (Hulme 2014). In a co-authored response, Stephan Lewandowsky and I argued that this was a counterproductive argument, given that those wishing to delay climate policy have focused on casting doubt on the consensus (Lewandowsky and Cook 2014). The entire point of reinforcing that the science on human-caused warming was settled was to move us to beyond the science to where the debate should be happening - what should we do about it?Attacks on our research continue unabated. Those wishing to delay climate action will continue to cast doubt on the scientific consensus. At the same time, climate scientists and science communicators will continue to debate the most effective ways to communicate the realities of climate change. And social scientists continue to research the role of consensus messaging. I am part of this research strand, exploring the psychology of consensus and the impact of misinformation on public perceptions.ReferencesAltmetric. 2014. Altmetric 2014 [cited Mar 1 2014].Anderegg, W. R., J. W. Prall, J. Harold, and S. H. Schneider. 2010. "Expert credibility in climate change." Proc Natl Acad Sci U S A no. 107 (27):12107-9. doi: 10.1073/pnas.1003187107.Bolsen, T., T.J. Leeper, and M.A. Shapiro. 2013. "Doing What Others Do: Norms, Science, and Collective Action on Global Warming." American Politics Research. doi: 10.1177/1532673X13484173Cook, J. 2014. Be part of a landmark citizen science paper on consensus. Skeptical Science 2013a [cited Mar 1 2014]. Available from http://www.skepticalscience.com/Be-part-of-landmark-citizen-science-paper-on-consensus.html.Cook, J. 2014. Communicating Scientific Consensus: John Cook responds. The Cultural Cognition Project at Yale Law School 2013b [cited Mar 1 2014]. Available from http://www.culturalcognition.net/john-cook-on-communicating-con/Cook, J., D. Nuccitelli, S.A. Green, M. Richardson, B. Winkler, R. Painting, R. Way, P. Jacobs, and A. Skuce. 2013. "Quantifying the consensus on anthropogenic global warming in the scientific literature." Environmental Research Letters no. 8 (2) (024024). doi: 10.1088/1748-9326/8/2/024024.Doran, Peter T., and Maggie Kendall Zimmerman. 2009. "Examining the Scientific Consensus on Climate Change." Eos, Transactions American Geophysical Union no. 90 (3):22-23. doi: 10.1029/2009EO030002.Elsasser, Shaun W., and Riley E. Dunlap. 2013. "Leading Voices in the Denier Choir: Conservative Columnists' Dismissal of Global Warming and Denigration of Climate Science." American Behavioral Scientist no. 57 (6):754-776. doi: 10.1177/0002764212469800.Forum, Climate Change National. 2014. Quantifying the consensus on anthropogenic global warming in the scientific literature" by John Cook et al: 97% of climate scientists agree that humans are causing global warming. Climate Change National Forum 2014 [cited Mar 1 2014]. Available from http://climatechangenationalforum.org/quantifying-the-consensus-on-anthropogenic-global-warming-in-the-scientific-literature-by-dr-john-cook-et-al-97-1-agree-that-humans-are-causing-global-warming/Hulme, M. 2014. Science can't settle what should be done about climate change. The Conversation 2014 [cited Mar 1 2014]. Available from http://theconversation.com/science-cant-settle-what-should-be-done-about-climate-change-22727Kahan, D. 2014. Annual "new study" finds 97% of climate scientists believe in man-made climate change; public consensus sure to follow once news gets out. The Cultural Cognition Project at Yale Law School 2013 [cited Mar 1 2014]. Available from http://www.culturalcognition.net/blog/2013/5/17/annual-new-study-finds-97-of-climate-scientists-believe-in-m.htmlKotcher, J., T. Meyers, E. Maibach, and A. Leiserowitz. 2014. Correcting misperceptions about the scientific consensus on climate change: Exploring the role of providing an explanation for the erroneous belief. In The International Communication Association.Lacatena, B. 2014. A Hack By Any Other Name - Part 1. 2014 [cited Feb 27 2014]. Available from http://www.skepticalscience.com/hack-2012-1.htmlLewandowsky, S., and J. Cook. 2014. Establishing consensus is vital for climate action. The Conversation 2014 [cited Mar 1 2014]. Available from https://theconversation.com/establishing-consensus-is-vital-for-climate-action-22861Lewandowsky, Stephan, Gilles E. Gignac, and Samuel Vaughan. 2013. "The pivotal role of perceived scientific consensus in acceptance of science." Nature Clim. Change no. 3 (4):399-404. doi: http://www.nature.com/nclimate/journal/v3/n4/abs/nclimate1720.html - supplementary-information.Obama, B. 2013. Remarks by the President on Climate Change. The White House 2013a2013]. Available from http://www.whitehouse.gov/the-press-office/2013/06/25/remarks-president-climate-changeObama, B. Twitter Post 2013b. Available from https://twitter.com/BarackObama/status/335089477296988160.Oreskes, N. 2004. "Beyond the ivory tower. The scientific consensus on climate change." Science no. 306 (5702):1686. doi: 10.1126/science.1103618.Oreskes, N. 2010. My facts are better than your facts,: spreading good news about global warming. Edited by M.S. Morgan and P. Howlett, How do facts travel? Cambridge: Cambridge Univesity Press.Pew. 2012. More Say There Is Solid Evidence of Global Warming. Pew Research Center for the People & the Press.

This article was originally published at The Winnower and is the backstory behind Edgar, R. C. 2004. "MUSCLE: multiple sequence alignment with high accuracy and high throughput." Nucleic Acids Res no. 32 (5):1792-7. doi: 10.1093/nar/gkh340.First off I should say that my career path was not planned or plannable, so you can't follow in my footsteps except to go where there is no path.From a young age (7 years old), my vocation was to be a scientist, and I quickly settled on theoretical physics because physics is everything - any explanation of the natural world eventually bottoms out in the laws of physics (yeah, I'm a hard-core reductionist and proud of it). At high school and college, I was pretty good at math and physics, but once I started my PhD I got a shock because I discovered that many people were much smarter than me. A few months after starting my postdoc I concluded that I would always be a mediocre physicist and would never make much of a contribution, so I quit. That was very difficult because physics had been my life's dream, but in retrospect physics was the wrong field for me; I'm more of a natural programmer than a natural mathematician. Like many failed physicists in the mid-80s, I got into the software business and ended up starting my own company in San Francisco. I sold the business to Intel in 1999, and in 2001 I was burned out and quit with no idea what to do next.I didn't want to start another business, which is what most entrepreneurs do, and I didn't want to retire and play golf (or more likely tennis). I knew about the human genome project and the importance of software algorithms - it was very appealing to me as a software guy that you could do important things just with strings of letters without knowing all that tedious biochemistry. Biology=strcmp (). So while it didn't occur to me that I might actively work on that stuff, I thought it would be fun to learn something about it, and crashed a seminar at UC Berkeley and met a newly-hired professor named Kimmen Sjolander. She was looking for students to help her code up some algorithms for summer research experience, and I volunteered. The result was a multiple alignment program called SATCHMO (Edgar and Sjolander 2003), which worked pretty well but was no better than CLUSTALW according to BALIBASE, which was the only available benchmark at that time.Being a competitive guy, I was dissatisfied with that, and started a systematic project to figure out what did and didn't work in multiple alignment algorithms, because it seemed to me that the programs had many ideas in them, but it wasn't clear which of the ideas helped or hurt the results. At that time, Gotoh's PRRx had the best accuracy (except maybe for T-Coffee; I don't remember) but the code was unfriendly and slow and hardly anybody used it. So I started from the PRRx algorithm and systematically varied the elements of the algorithm with all the alternatives I could find in the literature or dream up myself. That was the line of work that led to MUSCLE (Edgar 2004). (As an aside, the MAFFT people followed a similar strategy independently, and it's remarkable how much the original MAFFT and MUSCLE v1 resembled each other. MAFFT got published first and was better than my first attempt; I had to "borrow' a couple of their ideas in order to do better than MAFFT (Katoh et al. 2002). For example, we both had k-mer counting to build the first tree, but I was using 3-mers in the usual alphabet and they used 6-mers in a compressed alphabet, which gave slightly better results.So MUSCLE was a result of some financial independence I had after selling my business. I've supported myself and my research from my savings for the past decade. You can think of me as unemployed, independent and/or a gentleman scholar of modest means (like, say, my fellow-countryman Charles Darwin). It's hard to know how people have perceived my unconventional status. Everyone feels misunderstood and dismissed by peers sometimes (reviewers, editors, conference organizers), so it's impossible to say whether I would have been better accepted if I had a conventional affiliation. MUSCLE has been helpful because many people have heard of it. That gave me some street cred early on, and surely helped open other doors.I don't see any lessons here to help young scientists, so in an attempt to avoid becoming a bad influence I will defer to a more accomplished scientist, Richard Hamming, who has much better advice in his lecture, "You and Your Research" (Hamming 1986)ReferencesEdgar, R. C. 2004. "MUSCLE: multiple sequence alignment with high accuracy and high throughput." Nucleic Acids Res no. 32 (5):1792-7. doi: 10.1093/nar/gkh340.Edgar, R. C., and K. Sjolander. 2003. "SATCHMO: sequence alignment and tree construction using hidden Markov models." Bioinformatics no. 19 (11):1404-11. doi: 10.1093/bioinformatics/btg158.Hamming, R. You and Your Research 1986. Available from http://cmp.felk.cvut.cz/cmp/teaching/YouAndYourResearch.pdf.Katoh, K., K. Misawa, K. Kuma, and T. Miyata. 2002. "MAFFT: a novel method for rapid multiple sequence alignment based on fast Fourier transform." Nucleic Acids Res no. 30 (14):3059-66. doi: 10.1093/nar/gkf436.

This article was originally published at The Winnower and is the backstory behind Rozas, J., J.C. Sánchez-DelBarrio, X. Messeguer, and R. Rozas. 2003. "DnaSP, DNA polymorphism analyses by the coalescent and other methods." Bioinformatics no. 19 (18):2496-2497. doi: 10.1093/bioinformatics/btg359.The paper describes the version 4 of the DnaSP (DNA Sequence Polymorphism) software, a bioinformatics tool used to analyze and interpret DNA sequence polymorphism data (Rozas et al. 2003). The software implements a number of analytical methods, some of them developed by our own research group. Formally, the project began in 1994, but it has a history going back a few years ago, during my PhD studies (University of Barcelona, under the supervision of Professor M. Aguade) and, later during my Postdoc (Harvard University; at the laboratory of Professor R. C. Lewontin). At this time, I was performing some DNA variation analysis and, importantly, there was not any convenient software for the data analyses of such early studies. Thus, I started the development of DnaSP mainly because at that time there was no software to conduct such analyses.The analysis of DNA sequence polymorphisms and SNPs (single nucleotide polymorphisms) provides valuable insights into our understanding of the evolutionary meaning of DNA polymorphisms. For instance, both demographic (such as population expansions) or selective events, left distinctive molecular hallmarks on the patterns and levels of within-species DNA sequence data. Therefore, the analyses and interpretation of DNA polymorphism data can inform about, for example, past population growth or decline events, gene flow among populations, or the impact of natural selection (Begun et al. 2007, Nielsen 2005, Rosenberg and Nordborg 2002).Determining the impact of natural selection in molecular evolution and in adaptation is a fundamental question in evolutionary biology and evolutionary genomics. In this context, the detection of both positive (adaptive or Darwinian) and negative (purifying) selection is of great interest. Besides the deep conceptual implications, it might provide insights into our understanding of the biological function of genes and genomic regions. Indeed, positive selection promotes advantageous DNA sequence changes, which are ultimately responsible for evolutionary adaptations and novelties. The characterization of genomic regions shaped by this kind of selection has profound implications not only in evolutionary biology but also in our understanding of gene function. Negative selection, on the other hand, purges deleterious mutations favoring, therefore, the conservation of DNA sequences. Hence, these evolutionary conserved regions are, likely, to be functionally important and might uncover novel unknown biological functions. Moreover, DNA polymorphism data are very valuable powerful tools (as molecular markers) in a wide range of disciplines such as biomedicine, animal and plant breeding, conservation genetics, epidemiology genetics, or forensics.DnaSP is industry-standard software to compute population genetics analyses including coalescent-based methods, the state-of-the-art framework to analyze DNA sequence polymorphism data (Rozas et al. 2003, Rosenberg and Nordborg 2002, Librado and Rozas 2009, Wakeley 2009). DnaSP is a multi-purpose software package that allows conducting exhaustive analysis using a very friendly graphical user interface (GUI). The main features of the software are as follows: (i) accommodates large data sets; (ii) computes many population genetic statistics describing the level and patterns of DNA polymorphism within and between populations; (iii) conducts computer simulation coalescent-based tests; (iv) generates graphical outputs rendering the information readily understandable sequence polymorphism and SNP data.Although DnaSP version 4 was published in 2003, the first article (describing the first version of the software) was sent to publish to the scientific journal CABIOS (Computer Applications in Biosciences; currently named as Bioinformatics) in 1995, and it was quickly accepted (Rozas and Rozas 1995). Indeed, the other publications were also quickly accepted and published, including the most recent one (Librado and Rozas 2009). We think that this was because a combination of factors. First, it describes software for DNA polymorphism data analysis, and when the first version was released, there was no software of its kind. Second, it allows the computation of powerful and comprehensive state-of-the-art statistics and methods. Third, it provides a very easy to use GUI and with a learning curve that was extremely short. Fourth, the software can read a number of format data files, so it was very easy to input the data to the software to proceed with the analyses. Finally, the software provides an exhaustive help both to explain how to perform the analyses but also their scientific basis.ReferencesBegun, D.J., A.K. Holloway, K. Stevens, L.W Hillier, Y. Poh, M.W. Hahn, P.M. Nista, C.D Jones, A.D. Kern, C.N. Dewey, L. Pachter, E. Myers, and C.H. Langley. 2007. "Population Genomics: Whole-Genome Analysis of Polymorphism and Divergence in Drosophila simulans." PLoS Biol no. 5 (11):e310. doi: 10.1371/journal.pbio.0050310.Librado, P., and J. Rozas. 2009. "DnaSP v5: a software for comprehensive analysis of DNA polymorphism data." Bioinformatics no. 25 (11):1451-1452. doi: 10.1093/bioinformatics/btp187.Nielsen, R. 2005. "Molecular Signatures of Natural Selection." Annual Review of Genetics no. 39 (1):197-218. doi: 10.1146/annurev.genet.39.073003.112420.Rosenberg, N.A., and M. Nordborg. 2002. "Genealogical trees, coalescent theory and the analysis of genetic polymorphisms." Nat Rev Genet no. 3 (5):380-390. doi: 10.1038/nrg795.Rozas, J., and R. Rozas. 1995. "DnaSP, DNA sequence polymorphism: an interactive program for estimating population genetics parameters from DNA sequence data." Computer applications in the biosciences : CABIOS no. 11 (6):621-625. doi: 10.1093/bioinformatics/11.6.621.Rozas, J., J.C. Sanchez-DelBarrio, X. Messeguer, and R. Rozas. 2003. "DnaSP, DNA polymorphism analyses by the coalescent and other methods." Bioinformatics no. 19 (18):2496-2497. doi: 10.1093/bioinformatics/btg359.Wakeley, J. 2009. Coalescent Theory. An Introduction. Greenwood Village: Roberts and Company Publishers.

This article was originally published at The Winnower and is the backstory behind Peters, D.P., and S.J. Ceci. 1982. "Peer-review practices of psychological journals: The fate of published articles, submitted again." Behavioral and Brain Sciences no. 5 (02):187-195. doi:10.1017/S0140525X00011183.How the project was initiated. In 1978, one of us (SJC) was a newly hired assistant professor of psychology at the University of North Dakota, and the other one us (DPP) was in his fourth year as an assistant professor, one year away from the all-important tenure decision. One day at lunch with several colleagues, a discussion took place about bias in publishing. A senior colleague opined that a recent publication in a prestigious psychology journal would never have been published if the author had not been an eminent researcher from Harvard. He argued that if he had submitted the same manuscript for publication, it would have been rejected in a New York minute. We were not so sure. This individual's own difficulties getting his work published colored his view of science and prompted his conspiracy feelings.

This article was originally published on The Winnower and is the backstory behind Nittono, H., Fukushima, M., Yano, A., & Moriya, H. (2012). The power of kawaii: Viewing cute images promotes a careful behavior and narrows attentional focus, PLoS ONE, 7(9), e46362. doi:10.1371/journal.pone.0046362In this paper, I reported three experiments to demonstrate that people who viewed pictures of kittens and puppies, as compared to those who viewed pictures of adult animals, perform better on subsequent tasks that require dexterity and carefulness and tend to focus on the finer details of an object [1]. To be honest, I am not a pet lover. In more than forty years of life, my family and I have never had any kittens and puppies of our own. I have been a researcher in psychophysiology for twenty years, measuring activities in the brain and the autonomic nervous system in connection with various psychological phenomena. I had never dreamt of doing this kind of "pop" research up until a few years ago when a third-year student of mine asked me if she could study human reactions to kawaii items as the theme of her diploma thesis.Kawaii is one of the most popular words in contemporary Japan. Roughly speaking, it means "cute," "lovely," "pretty," "adorable," etc. This word is frequently uttered to express the speaker's favorable evaluation toward an object or person. Because it is used too often, sociologists and journalists have discussed for decades that kawaii is a key concept to understand the characteristics of modern Japanese people and culture. Nevertheless, many believed that it was out of the scope of science. "It is difficult," I said, "but if you really want to do so, try it." While she recorded psychophysiological responses to human and animal baby pictures [2], I carefully read through many articles regarding kawaii, from "The Pillow Book," an essay written by a Japanese court lady around the year 1000, to the most recent fashion magazines. As a psychologist, I was curious why people love kawaii so much and thought that there should be a psychological factor that reinforces the behavior of admiring and possessing kawaii items. At that point I was confident that, when done well, this line of research would attract much public attention, at least inside Japan.Shortly after we started this investigation, an interesting psychology paper was published in the US. The authors argued that viewing pictures of baby animals promotes behavioral carefulness [3]. Because the results initially sounded odd but intriguing, I tried to replicate and extend the findings in a Japanese sample with support of three other students of mine. This is how and why I started the experiments reported in the paper in question.As many other projects were going on in parallel, it took about two years to complete three experiments. The paper was first submitted to Psychological Science, which is one of the most highly reputed journals that publish many exciting psychology papers. However, it was rejected immediately without being sent out to reviewers, because the paper made "a useful but incremental contribution to the literature" rather than reporting "groundbreaking findings." Needless to say, I had been aware that the study lacked true originality and was, in essence, a replication. On the other hand, I was pretty sure that this topic would attract public attention, so we decided to submit it to an open access journal, PLoS ONE. Before resubmission, however, I added a twist. Originally, the paper was titled "The power of cuteness," which was more comprehensible for an international audience. However, I deliberately changed it to "The power of kawaii," with the Japanese public in mind. I also inserted a short paragraph at the beginning of the paper to introduce what kawaii is and what it means in Japanese culture.After the second revision, it was published in September 2012. Fortunately, the editor and reviewers never asked me to change the title to a "plainer" one. Therefore, it became the first peer-reviewed scientific paper with the word kawaii in the title among more than 23 million citations in PubMed. Before publication, for the first time in my life I asked a public information officer from my university to put out a media release.The research was positively received. On the day of publication and the following several days, I was interviewed by a dozen of Japanese newspapers, radio hosts, and TV personalities. Furthermore, over seventy international news and blogs featured our findings, which surpassed my predictions. There were remarkable differences in the focus of interest between Japanese and international news media. While Japanese media mainly focused on the idea of scientific research on kawaii, other media passed over this aspect and just focused on the finding that viewing pictures of kittens and puppies may benefit work performance.What enabled this research study to generate so much attention? It was partly due to luck, of course. With hindsight, however, there seem to be several factors that increased the likelihood of an article gaining wider attention in society. First, the study originated from a layperson's question. Second, the topic was well-framed in a local cultural background. Third, the finding itself was relevant to a wider, international audience for different reasons than those that pertained to the original culture. Fourth, it was published in a popular open access journal that is easily accessible by everyone via the Internet. Last but not least, what I learned from this experience is that, in order to publish a paper that is influential amongst the general public, it is important to conduct not only a groundbreaking study but also a study that replicates previous findings and extends them in a proper direction that many people are interested in learning more about.REFERENCES1. Nittono, H., Fukushima, M., Yano, A., & Moriya, H. (2012). The power of kawaii: Viewing cute images promotes a careful behavior and narrows attentional focus, PLoS ONE, 7 (9), e46362. doi:10.1371/journal.pone.00463622. Nittono, H., & Tanaka, K. (2010). Psychophysiological responses to kawaii (cute) visual images [abstract]. International Journal of Psychophysiology, 77, 268-269. doi:10.1016/j.ijpsycho.2010.06.1073. Sherman, G. D., Haidt, J., & Coan, J. A. (2009). Viewing cute images increases behavioral carefulness. Emotion, 9, 282-286. doi: 10.1037/a0014904

This document was originally published at The Winnower and is the backstory behind Hart, V., P. Novakova, E. P. Malkemper, S. Begall, V. Hanzal, M. Jezek, T. Kusta, V. Nemcova, J. Adamkova, K. Benediktova, J. Cerveny, and H. Burda. 2013. "Dogs are sensitive to small variations of the Earth's magnetic field." Front Zool no. 10 (1):80. doi: 10.1186/1742-9994-10-80.Almost half a century ago, the first firm evidence for magnetoreception (i.e. the ability to sense the geomagnetic field) in the robin was provided, representing the first evidence for magnetoreception in a vertebrate (Wiltschko and Merkel 1966); a quarter century later Burda et al. (Burda et al. 1990) demonstrated magnetoreception for the first time in the laboratory in African mole-rats, the first demonstration in mammals. Over the last 25 years we have revisited the question of magnetoreception in mole-rats and made some spectacular discoveries (Marhold, Wiltschko, and Burda 1997, Nemec et al. 2001, Thalau et al. 2006, Wegner, Begall, and Burda 2006, Burger et al. 2010). Subsequently, magnetoreception has been described also in the blind mole rat Spalax ehrenbergi supersp. (Kimchi and Terkel 2001, Kimchi, Etienne, and Terkel 2004, Marhold et al. 2000), the laboratory mouse Mus domesticus (Muheim et al. 2006), the Siberian hamster Phodopus sungorus (Deutschlander et al. 2003), and in the bats Eptesicus fucus (Holland et al. 2006, Holland et al. 2008) and Nyctalus plancyi (Wang et al. 2007). However, the question if large mammals also sense the geomagnetic field was left unanswered.How to study magnetoreception in large mammals Some mammals, such as arctic caribous, North American bisons, African elephants and gnus, and Asian saiga antelopes, and, above all, diverse cetacean species are famous for their long range migrations, which require superb navigational capacities. Based on these capacities and on correlations between navigation failures and local disturbances of the geomagnetic field we may conclude that navigation capacities are based on magnetoreception (Klinowska 1985, Kirschvink, Dizon, and Westphal 1986). Apart from such indirect evidence there are also anecdotal reports on homing abilities of dogs, cats, horses etc. (just remember "Lassie Come Home"). Systematic research on navigational capabilities using homing paradigms is technically difficult. Besides that, there could be plenty of non-navigational reasons why displaced mammals do not return home. Conditioning (training) on magnetic stimuli in large animals is methodically difficult. The animals might switch on the magnetic compass sense only in certain contexts, e.g. when being outdoors, in unfamiliar surroundings, where they cannot rely on other senses, but not under laboratory conditions.However, looking for spontaneous expressions of magnetosensitivity, such as magnetic alignment is an easy and effective method to scan for magnetoreception in animals. Animals (and also humans) frequently adopt orientation in space which is not random (Begall et al. 2013): e.g. when sun basking they align perpendicular to solar rays (thermic alignment), trout in the brook aligns against the stream (rheo-alignment), a scenting dog aligns against the wind (anemo-alignment), the audience turns towards the speaker (alignment towards an attractor), grazing sheep on a slope wander along the contours (isocline-alignment). Apparently alignment has some advantages - it is comfortable, saves energy, or provides access to food, oxygen or information. And now imagine that there is no slope, no attractor, no wind and the sun is not visible. Will the orientation (body axis, heading) of a resting animal be random or might the animal prefer to adopt some particular compass direction? If we find that animals prefer a particular compass direction under these unconstrained conditions, we may hypothesize that such a non-random orientation represents a case of magnetic alignment, implying magnetoreception. Subsequently, we may test this hypothesis by changing the magnetic conditions and observing whether the orientation also changes. Scanning for cases of magnetic alignment is thus a suitable experimental paradigm for magnetobiological research and might provide new insights into meaning of magnetoreception in animals' everyday life. Why do animals do it? Which information do they get from the magnetic field? Can they save energy (get more comfort) when magnetically aligned? How do they sense the magnetic field?For over six years we have been following this research program and have reported magnetic alignment in grazing and resting cattle, roe deer and red deer (Begall et al. 2008) and disturbance of this alignment through extremely low frequency magnetic fields (ELF MF) produced by high voltage power lines (Burda et al. 2009). These papers raised high resonance all over the world. Skeptics did not like particularly the idea that ELF MF might influence the behavior (and welfare?) of animals (and human health?). They attacked our results (Hert et al. 2011), but we could show (Begall et al. 2011) that their study was heavily flawed and cleaned of errors it actually fully confirmed our original findings. Our results were subsequently confirmed in an independent study by Slabý et al. (Slaby, Tomanova, and Vacha 2013). In other studies we found magnetic alignment in hunting foxes (Cerveny et al. 2011), carps schooling in circular tubs (Hart et al. 2012), and mallard ducks landing on water (Hart et al. 2013). We concluded that magnetic alignment might help animals to synchronize and coordinate movement in a group, organize cognitive (mental) maps, measure the distance and the slope, and maybe enhance (focus) selective sensory attention. Our model mammals (cattle, deer, fox) were, however, rather "unhandy" and unmanageable for further research. We envisaged that dogs represent ideal model animals for further research, since they are widespread across the world and are highly receptive to learning.How to study magnetoreception in dogsOur established team of sensory ecologists from Essen, Germany, and wildlife zoologists from Prague, Czech Republic, asked dog owners (relatives, friends and students) to measure, in analogy with our previous research on cattle and deer, body orientation of feeding and sleeping dogs. We soon realized, however, that body orientation during these activities is influenced, consciously or unconsciously by the master, by house geometry etc. Besides that, the geomagnetic field inside buildings is mostly disturbed, and it would require many cooperators to get sufficient sample sizes. We noticed, however, that dogs defecating and urinating during a walk (actually they are marking their territories) are very selective about where to mark. We can prevent them to mark somewhere but we can hardly make them do the job when and where we wish. Well known, yet enigmatic and unexplained (though there are many ideas) is the circling (some) dogs sometimes perform before urinating or defecating. Every dog defecates on average once and urinates (marks) many times a day; keepers go with their dogs at least three times a day to a walk; if every walk (in the morning, afternoon, and evening) would follow another route, measurements of orientation of particular dogs cannot be considered pseudoreplications, as they would be done always at a different time, at different place, under somewhat different environmental conditions. We standardized the protocol, asked our volunteers to record compass direction of the body axis (heading) of defecating and urinating dogs, which would be unrestrained by walls and fences, i.e. in open field, off the leash, and outdoors (i.e. not at the home backyard or in the familiar garden). We recruited a team of almost 40 observers following 70 dogs of almost 40 breeds over two years, recording about 1,900 defecation and 5,600 urination events. After we closed sampling, we ran circular statistical analyses and were disappointed to see that there was no common preference for any particular compass direction. We analyzed the data with respect to sex, age, breed, and body mass of the dog, time of the day and month but no clear picture emerged. The subsamples, i.e. the numbers of records collected for particular dogs were not even. One dog, a male borzoi named Diadem, was recorded on a regular basis over a long time providing about 2,500 measurements of body orientation. Diadem's data showed to be a gold mine. We noticed that on some days and during some walks Diadem showed perfect compass preference (in his case for Northwest), while during some other walks and on some particular dates his orientation during marking was fully random. It was just by chance that, when analyzing the data, we heard a radio report about geomagnetic storms and the idea was born to check the "space weather" on those days when Diadem was apparently "disoriented". A comparison between "quiet" and "stormy" days confirmed that alignment could be correlated to "magnetic weather". In subsequent weeks and months we spent many hours of leisure time looking for different indices of magnetic weather at time and place of the particular walks and checking which of them correlate best with the "alignment achievement" of the dogs. We finally found that the best predictor of alignment was the speed (rate, slope) of changes in the magnetic declination at the time interval 30 minutes before the dog was urinating or defecating. This single factor had the power to extract a highly significant alignment from a random distribution.Most volunteers were aware of our previous findings and an unconscious bias in measurements could not be excluded. (Meanwhile we developed smart phone apps to measure the compass direction blindly, i.e., without showing the measured value.) None of us or of the volunteers who measured their dogs knew, however, the "magnetic weather" when taking or analyzing the measurements. Thus, the study was indeed fully blind.Why do the dogs do it?The dog has a mental map of its home range or creates such a map whenever being in an unfamiliar region. Our hypothesis is that when the dog marks or defecates, it uses these short stops to store the location ("coordinates") of marking cornerstones (as landmarks) in its memory. Doing so and/or calibrating its magnetic compass is probably easier for it when being aligned with the magnetic field. (When we read a map and use a compass outdoors we also align according to some directions and landmarks). Apparently, the dog needs calm "magnetic weather" to do so.The publication and its impactWe were quite excited about our findings as this was the first evidence for a) magnetoreception in dogs, b) sensitivity to such small changes of the geomagnetic field in a mammal, c) sensing small changes in polarity (rather than in intensity) of the magnetic field in an animal. Knowing that dogs can sense magnetic fields opens up new possibilities for investigation of principles and mechanisms of magnetoreception. Knowing that animals may sense small oscillations in the geomagnetic field and react with a behavioral change calls for considering this factor when studying and analyzing many aspects of animal behavior.Of course, we wanted to publish our results in a top multidisciplinary journal. Yet all the journals returned the manuscript immediately - as being of only minor interest for their general readerships. We assume that the editors were afraid to become embarrassed by publishing a study dealing with something as trivial as defecation. Maybe they have misunderstood its message. We are very grateful to Frontiers in Zoology, a top journal in zoology, for letting the manuscript to be peer reviewed. After passing the very quick but also very thorough review process, the paper was published during Christmas holidays in 2013. Still before our universities published a press release and before we made any advertisement for it, a report about it was published in the Journal of Improbable Research. Within one week after publication, the paper achieved an altmetric score of 1,715 and was accessed about 740,000 times. (The score rose to 2,854 and the number of accesses to almost 880,000 three months later.) We were contacted by many journals and newspapers, internet sites, radio and TV news, professionals and interested laymen. Many published reports were based on tertiary sources and most of the "reporters" have never read the original article, although it is freely accessible in the internet. The message of our article was reduced to the statement that dogs prefer to face northwards when defecating. (Something we never claimed.) Many applauded while others berated. We were often rebuked for wasting tax money for nonsense research. People claiming this have perhaps never deliberated on possible costs of such studies. This research on dogs was done in our leisure time - we walked with our dogs in the same way as millions of people do all over the world - but we took also compass, paper and pencil, and in the evening when others were watching TV, surfing in the web or going in for their hobbies, we analyzed our results. So - it was our hobby and we had fun doing it. We have experienced immense reactions upon our research already before, when we described magnetic alignment in cattle. Our studies dealing with other animals aroused much less interest. Based on our experience we were not surprised that our article on dogs made quite a splash. Many dog owners assume they "know" everything about dogs and are qualified to comment. Very few people have pet deer, foxes, mole-rats or shrews, so our work on these animals did not challenge them personally.A study that makes people laugh and then they start to think The Journal of Improbable Research which started this hype is dedicated to research that makes people laugh and then think. We believe that this will be also the fate of our paper. Indeed, finding that dogs can sense and recognize small fluctuations of the geomagnetic field provides us with lots of stuff to think seriously about.ReferencesBegall, S., H. Burda, J. Cerveny, O. Gerter, J. Neef-Weisse, and P. Nemec. 2011. "Further support for the alignment of cattle along magnetic field lines: reply to Hert et al." J Comp Physiol A no. 197 (12):1127-1133. doi: 10.1007/s00359-011-0674-1.Begall, S., J. Cerveny, J. Neef, O. Vojtech, and H. Burda. 2008. "Magnetic alignment in grazing and resting cattle and deer." Proc Natl Acad Sci U S A no. 105 (36):13451-5. doi: 10.1073/pnas.0803650105.Begall, S., E. P. Malkemper, J. Cerveny, P. Nemec, and H. Burda. 2013. "Magnetic alignment in mammals and other animals." Mammalian Biology no. 78 (1):10-20. doi: 10.1016/j.mambio.2012.05.005.Burda, H., S. Begall, J. Cerveny, J. Neef, and P. Nemec. 2009. "Extremely low-frequency electromagnetic fields disrupt magnetic alignment of ruminants." Proc Natl Acad Sci U S A no. 106 (14):5708-13. doi: 10.1073/pnas.0811194106.Burda, H., S. Marhold, T. Westenberger, W. Wiltschko, and R. Wiltschko. 1990. "Magnetic compass orientation in the subterranean rodent Cryptomys hottentotus (Bathyergidae, Rodentia)." Experientia (46):528-530.Burger, T., M. Lucova, R. Moritz, H. OelschlŠger, R. Druga, H. Burda, W. Wiltschko, R. Wiltschko, and P. Nemec. 2010. "Changing and shielded magnetic fields suppress c-Fos expression in the navigation circuit: input from the magnetosensory system contributes to the internal representation of space in a subterranean rodent." Journal of The Royal Society Interface no. 7 (50):1275-1292. doi: 10.1098/rsif.2009.0551.Cerveny, J., S. Begall, P. Koubek, P. Novakova, and H. Burda. 2011. "Directional preference may enhance hunting accuracy in foraging foxes." Biology Letters no. 7 (3):355-357. doi: 10.1098/rsbl.2010.1145.Deutschlander, M., M. Freake, C. Borland, J.B. Phillips, R.C. Madden, L. Anderson, and B. Wilson. 2003. "Learned magnetic compass orientation by the Siberian hamster, Phodopus sungorus." Animal Behaviour no. 65 (4):779-786. doi: 10.1006/anbe.2003.2111.Hart, V., T. Kusta, P. Nemec, V. Blahova, M. Jezek, P. Novakova, S. Begall, J. Cerveny, V. Hanzal, E. P. Malkemper, K. Stipek, C. Vole, and H. Burda. 2012. "Magnetic alignment in carps: evidence from the Czech christmas fish market." PLoS One no. 7 (12):e51100. doi: 10.1371/journal.pone.0051100.Hart, V., E. P. Malkemper, T. Kusta, S. Begall, P. Novakova, V. Hanzal, L. Pleskac, M. Jezek, R. Policht, V. Husinec, J. Cerveny, and H. Burda. 2013. "Directional compass preference for landing in water birds." Front Zool no. 10 (1):38. doi: 10.1186/1742-9994-10-38.Hert, J., L. Jelinek, L. Pekarek, and A. Pavlicek. 2011. "No alignment of cattle along geomagnetic field lines found." Journal of Comparative Physiology A no. 197 (6):677-682. doi: 10.1007/s00359-011-0628-7.Holland, R.A., J.L. Kirschvink, T.G. Doak, and M. Wikelski. 2008. "Bats use magnetite to detect the Earth's magnetic field." PLoS ONE no. 3 (2):e1676. doi: 10.1371/journal.pone.0001676.Holland, R.A., K. Thorup, J.M. Vonhof, W.W Cochran, and M Wikelski. 2006. "Navigation: Bat orientation using Earth's magnetic field." Nature no. 444 (7120):702-702. doi: 10.1038/444702a.Kimchi, T., A. S. Etienne, and J. Terkel. 2004. "A subterranean mammal uses the magnetic compass for path integration." Proc Natl Acad Sci U S A no. 101 (4):1105-9. doi: 10.1073/pnas.0307560100.Kimchi, T., and J. Terkel. 2001. "Magnetic compass orientation in the blind mole rat Spalax ehrenbergi." J. Exp. Biol. (204):751-758.Kirschvink, J.L., A.E. Dizon, and J.A. Westphal. 1986. "Evidence from strandings for geomagnetic sensitivity in cetaceans." J. Exp. Biol. (120):1-24.Klinowska, M. 1985. "Cetacean stranding sites relate to geomagnetic topography." Aquatic Mammals (1):27-32.Marhold, S., A. Beiles, H. Burda, and E. Nevo. 2000. "Spontaneous directional preference in a subterranean rodent, the blind mole-rat, Spalax ehrenbergi." Folia Zool. (49):7-18.Marhold, S., W. Wiltschko, and H. Burda. 1997. "A magnetic polarity compass for direction finding in a subterranean mammal." Naturwissenschaften (84):421-423. doi: 10.1007/s001140050422.Muheim, R., N. M. Edgar, K. A. Sloan, and J. B. Phillips. 2006. "Magnetic compass orientation in C57BL/6J mice." Learn Behav no. 34 (4):366-73.Nemec, P., J. Altmann, S. Marhold, H. Burda, and H. Oelschlaeger. 2001. "Neuroanatomy of magnetoreception: The superior colliculus involved in magnetic orientation in a mammal." Science no. 294 (5541):366-368. doi: 10.1126/science.1063351.Slaby, P., K. Tomanova, and M. Vacha. 2013. "Cattle on pastures do align along the North-South axis, but the alignment depends on herd density." Journal of Comparative Physiology A no. 199 (8):695-701. doi: 10.1007/s00359-013-0827-5.Thalau, P., T. Ritz, H. Burda, R. E. Wegner, and R. Wiltschko. 2006. "The magnetic compass mechanisms of birds and rodents are based on different physical principles." J R Soc Interface no. 3 (9):583-7. doi: 10.1098/rsif.2006.0130.Wang, Y., Y. Pan, S. Parsons, M. Walker, and S. Zhang. 2007. "Bats respond to polarity of a magnetic field." Proc Biol Sci no. 274 (1627):2901-5. doi: 10.1098/rspb.2007.0904.Wegner, R. E., S. Begall, and H. Burda. 2006. "Magnetic compass in the cornea: local anaesthesia impairs orientation in a mammal." J Exp Biol no. 209 (Pt 23):4747-50. doi: 10.1242/jeb.02573.Wiltschko, W., and F.W. Merkel. 1966. "Orientierung zugunruhiger Rotkehlchen im statischen Magnetfeld." Verh. Dtsch. Zool. Ges. (59):362-367.

This article was originally published at The Winnower and is the back story behind Cole, T.J., M.C. Bellizzi, K.M. Flegal, and W.H. Dietz. 2000. "Establishing a standard definition for child overweight and obesity worldwide: international survey." BMJ no. 320 (7244):1240. doi: 10.1136/bmj.320.7244.1240.Obesity in humans came to be recognised as a serious public health concern during the 1960's, when several national and international obesity societies were founded. But the focus was initially on adults, and it took a further 30 years for the concern to extend to children. This was because obesity is to some extent an inter-generational condition, with fat parents begetting fat children, so it took time - more than a generation - for the levels of obesity in children to rise sufficiently to cause alarm. Once the danger was recognised, researchers from all over the world rushed to carry out prevalence studies to quantify the scale of the problem, and many such studies were published during this time.However there was a serious problem - nobody could agree on how to define child obesity. As a result each study used its own definition, with the unfortunate consequence that the methods were incompatible and the prevalence rates could not be compared across studies. This highlighted the need for a unified international definition that all could use, which would allow the effort being put into measuring child obesity prevalence to be properly harnessed.The International Obesity Task Force (IOTF) was another reflection of the rising concern about obesity. It was set up in 1994 by Philip James as a policy and advocacy think tank, and its initial action was to prepare what became the first scientific report on the global epidemic of obesity, published as a WHO expert technical report in 2000. This work highlighted the lack of a proper definition for child obesity, and a second IOTF initiative was to convene a workshop on child obesity in Dublin in June 1997 to address the deficiency. The purpose and summary of the workshop were later published (Dietz and Bellizzi 1999, Bellizzi and Dietz 1999). The primary outcome of the workshop was a definition of childhood obesity that appeared in the BMJ (Cole et al. 2000), a paper which has since been cited 5,391 times by Web of Science (an average of 350 citations per year) and 10,125 times by Google Scholar (as of 8 May 2014).Obesity is more complicated to diagnose in children than in adults, because children increase in height, weight and body fat naturally as they grow. The difficulty is to decide when the pattern of growth is excessive. So to define child obesity requires three components: a measure to define the amount of body fat; a reference population to define the distribution of the body fat measure by age and sex, and a centile cut-off to define the point in the body fat measure distribution corresponding to obesity. For historical reasons there were in fact two cut-offs, one for obesity and a less extreme one for overweight. The workshop agreed to use body mass index (BMI or weight/height 2) as the measure of body fat, an uncontroversial choice as it matched the usage in adults. However the other two components of the definition were novel: large nationally representative samples of children from Brazil, Great Britain, Hong Kong, the Netherlands, Singapore and USA (n = 192,727) that were used to establish the distribution of BMI internationally; and (my idea) centile cut-offs of BMI chosen to match the universally accepted adult cut-offs of BMI 25 (overweight) and BMI 30 (obesity) at age 18 years. This ensured that the child and adult definitions were consistent with each other, a notable advance on previous definitions.Centile charts are widely used to assess growth in paediatric practice, and previous obesity definitions had used the 85th and 95th centiles of BMI as overweight and obesity cut-offs. The shapes of the centile curves on the chart can be estimated with a statistical technique called the LMS (λ-μ-σ) method which I developed with Peter Green (Cole and Green 1992). LMS allows one to construct any required centiles, in particular those that pass through BMI 25 and 30 at age 18. The calculations were done for each of the six samples separately, and the cut-offs were then averaged.One indicator to justify the approach taken was to see how close to each other the centiles based on the separate samples were. It turned out that the Singapore cut-offs were a different shape from all the others, the Singaporean children being relatively fatter in early puberty (Cole et al. 2000), and the first version of the paper excluded the Singapore cut-offs on that basis. However the BMJ reviewers said they should be retained, and in retrospect this was a wise decision - the anomalous pattern of BMI change with age seen in the Singapore sample has since been seen in other datasets too.Another early intention of the authors was to develop cut-offs to identify underweight as well as overweight, based on the same BMI reference data. Child malnutrition is as big a public health problem as obesity in many parts of the world, and cut-offs based on BMI would have advantages over the weight-for-height cut-offs then in use. However it proved to be too ambitious at the time, and only in 2007 was the follow-up paper published (Cole et al. 2007). It has been cited 580 times to date. More recently the underlying methodology has been overhauled, and it is now possible to present the information in the form of a BMI growth chart (Cole and Lobstein 2012).It is interesting to speculate as to why the 2000 paper was such a runaway success. The first ingredient was its authors: Bill Dietz and Katherine Flegal were both well-known and respected obesity researchers from the US Centers for Disease Control and Prevention, which encouraged the paper's take-up in the USA. Mary Bellizzi was Secretary of the IOTF, and this emphasised the international nature of the definition. I provided the statistical framework on which the work was based. The second ingredient was the wide variety of countries providing data, giving reassurance to researchers that the cut-offs were indeed international and could be applied with confidence to children from their own country. By contrast the World Health Organization (WHO) international growth reference was at that time based on US data, and this was politically uncomfortable in parts of the world where the pattern of child growth was clearly different from that in the USA.A third ingredient was the way the cut-offs were constructed, linked to the widely used adult definition. This provided reassurance that the cut-offs were not just a statistical artefact but were based on existing practice. And finally the paper's publication in the BMJ, a general medical journal with very broad reach, ensured that it received widespread publicity from day one. The BMJ itself was impressed by the high citation rate of the paper, and used it as the example in a discussion piece about how the internet was altering the nature of literature citations (Delamothe 2002).Overall the paper's timeliness was the key - it addressed a problem that researchers worldwide were struggling with, while at the same time providing them with psychological and political reassurance that it was valid to use. This was confirmed by an exhaustive IOTF review of child obesity published in 2004 which quoted many studies with prevalence rates based on the definition (Lobstein, Baur, and Uauy 2004). As the Abstract Conclusion of the 2000 paper concisely summarised it,"The proposed cut off points, which are less arbitrary and more internationally based than current alternatives, should help to provide internationally comparable prevalence rates of overweight and obesity in children."ReferencesBellizzi, M. C., and W. H. Dietz. 1999. "Workshop on childhood obesity: summary of the discussion." American Journal of Clinical Nutrition no. 70 (1):173S-175S.Cole, T J, M C Bellizzi, K M Flegal, and W H Dietz. 2000. "Establishing a standard definition for child overweight and obesity: international survey." British Medical Journal no. 320 (7244):1240-3.Cole, T J, K M Flegal, D Nicholls, and A A Jackson. 2007. "Body mass index cut-offs to define thinness in children and adolescents: international survey." British Medical Journal no. 335:194-7.Cole, T. J., and P. J. Green. 1992. "Smoothing reference centile curves: the LMS method and penalized likelihood." Statistics in Medicine no. 11 (10):1305-19.Cole, T. J., and T. Lobstein. 2012. "Extended international (IOTF) body mass index cut-offs for thinness, overweight and obesity." Pediatr Obes no. 7 (4):284-94. doi: 10.1111/j.2047-6310.2012.00064.x.Delamothe, T. 2002. "Is that it? How online articles have changed over the past five years." BMJ no. 325:1475. doi: doi: 10.1136/bmj.325.7378.1475.Dietz, W. H., and M. C. Bellizzi. 1999. "Introduction: The use of body mass index to assess obesity in children." American Journal of Clinical Nutrition no. 70 (1):123S-125S.Lobstein, T, L Baur, and R Uauy. 2004. "Obesity in children and young people: a crisis in public health." Obesity Reviews no. 5 (Suppl 1):4-85.

This article was originally published at The Winnower and is the backstory behind Guindon, S., and O. Gascuel. 2003. "A simple, fast, and accurate algorithm to estimate large phylogenies by maximum likelihood." Syst Biol no. 52 (5):696-704.I started working on PhyML during my second year as a PhD student. The article describing the first part of my PhD thesis had just been published and I felt it was the right time to take some risk and try something which first seemed out of my depth: implementing a program that calculates the phylogenetic likelihood function. In 2002, only very few softwares were based on the likelihood principle. The calculation of this function appeared to me as a tough challenge, but the underlying algorithm (Felsenstein's prunning algorithm (Felsenstein 1981)) is beautiful and I was thus eager to test my programming skills on that nice problem. I was based in Montpellier, in the south of France, at that time, but my wife lived in Paris which means I was spending a lot of time away from the lab. This freedom gave me the opportunity to immerse myself completely in my task. I remember being in Paris, not far from the Sacre Coeur, crunching numbers and, for the first time, having my own program return the very same likelihood value as that produced by PAML (Yang 2007) and PHYLIP (Felsenstein 2005), the references in likelihood-based phylogenetic softwares. This felt like a very significant victory to me. I was hooked.I thus continued programming and tried to accommodate for larger data sets and apply more sophisticated parameter estimation techniques. It quickly appeared though that conventional algorithms would not allow me to analyze data sets with more than ~10 sequences. Other methods, that do not rely on the likelihood framework, could easily go up to ~100 sequences but lacked accuracy. A significant speed up in likelihood-based phylogenetic analyses was therefore in dire need. The core of my program relied on functions that would modify the current solution one step at a time, with each step applying the same operation to a new part of the phylogenetic tree. In order to save computing time, I decided to slightly modify that core and apply these multiple local operations all at the same time. Surprisingly, the results turned out to be very encouraging: the new algorithm was not only as accurate as the other likelihood-based softwares, it was also an order of magnitude faster. I remember then proudly showing the first results to Olivier Gascuel, my PhD supervisor. He was quite enthusiastic too and suggested further optimization strategies that significantly improved the algorithm. PhyML was born.Olivier then wrote most of the paper while I was running extensive simulations that would compare the speed and accuracy of PhyML to its competitors. We first sent the article to PNAS, which rejected it after only a couple of days. We then tried Systematic Biology and asked Bruce Rannala to act as associate editor. He and the reviewers did a wonderful job that contributed to substantially improve the original draft. The paper was finally published in October 2003 (Guindon and Gascuel 2003). I started receiving requests from the first users of PhyML asking me to add new options to the program. People seemed to like the software overall and the article started to get cited. We reached 1,000 citations in 2007 and 5,000 four to five years after that, which made that paper the second most-cited in the field of ecology and evolution in the 1998-2008 period.In my opinion, the key ingredients in this project were a bit of risk-taking at the beginning, a lot of efforts spent in programming (read: fixing bugs!) and a great student-supervisor relationship. We were luckly to come up with this new algorithm at a time where DNA sequencing was gathering considerable pace and people were therefore eager to speed up their phylogenetic analysis while keeping the same level of accuracy. So, good timing was therefore paramount too.ReferencesFelsenstein, J. 1981. "Evolutionary trees from DNA sequences: a maximum likelihood approach." J Mol Evol no. 17 (6):368-76. doi: 10.1007/BF01734359.Felsenstein, J. PHYLIP (Phylogeny Inference Package) version 3.6, Department of Genome Sciences, University of Washington, Seattle.Guindon, S., and O. Gascuel. 2003. "A simple, fast, and accurate algorithm to estimate large phylogenies by maximum likelihood." Syst Biol no. 52 (5):696-704. doi: 10.1080/10635150390235520.Yang, Z. 2007. "PAML 4: phylogenetic analysis by maximum likelihood." Mol Biol Evol no. 24 (8):1586-91. doi: 10.1093/molbev/msm088.