Cascais
Snapshot from Lisbon #4
Leave a reply
Category Archives: Uncategorized
Aside: Marissa Mayer, Yahoo!, and being pretty annoying
If you have one hour of your time sitting there for you, and going to beach, playing with your kids, drink something etc. does not seem a good option, I recommend this Business Insider's article on Marissa Mayer. Super-driven, clearly very smart, so far her Yahoo! tenure has been a huge success (but read how Alibaba must be partially responsible), but there is a but. Quite a few buts, actually. Would I enjoy working under her? Not at all. Always late, she tries to out-talk you, very often she cancels one-on-one meeting without previous notice and no justifications, makes people wait for her forever, no social skills, she says that raising baby is easy when she has an army of helpers, nannies, a $5 million penthouse, come on.
"Mayer had approximately 25 people reporting directly to her during her first year at Yahoo. In theory, she was keeping up with each of them in a regularly scheduled weekly meeting. In practice, she would go weeks without talking to people because she was so busy.
For a while, each of those 25 people thought that Mayer was just picking on them, individually. The people who had been at Yahoo before Mayer joined assumed that this meant she was going to fire them soon. The people Mayer had just hired into the company, including Reses and Savitt, were even more puzzled. Why had they been hired only to be ignored?"
I like analytical people, I don't enjoy that much the "slap in the back" attitude, especially when I am on the receiving side of slapping, I am all for talking shop up to 2am and getting back at it after 3 hours of sleep, but cancelling appointments like the other guy is worth neither your "I am sorry" nor some kind of explanation, having open office hours like you are an high school teacher, well, I think I'd last 10 minutes. A lot of people left Yahoo! after she got the CEO position, and apart from business and personality clashes inevitably happening at that level, a lot of decent people I believe simply couldn't put up with the disrespectful behavior and attitude.
I do not believe there are "success" lessons in the Mayer's story. Luck makes a huge difference in people's lives and thinking that the road to work or business success can be algorithmically explained or predicted is naive at best. Smart, hard working, brilliant people are on average (and there is a threshold effect for some of those traits, simply people who are not smart in a loose and broad sense are very unlikely to be successful, although we often associate smartness to quantitative and mathematical skills and intuition, and it is quite limiting) more likely to succeed than less smart, less brilliant and lazier people, but beyond that everything is really context-specific and hugely dependent on being at the right place at the right time. Looking at success stories backwards sometimes gives clues, a lot of times doing that is very misleading. The Success Equation is a good read here.
The article, although quite verbose at times, gives a solid and interesting picture of how business deals (including getting a new CEO on board at Yahoo!) are made at the highest level. Very interesting and highly recommended.
Snapshot from Lisbon #1
ESEB meeting recap
The ESEB meeting left me quite satisfied, although the proportion of talks I really liked was probably ~ 1 out of 5. I think that the submission of abstract 6 months in advance is partially responsible of some sub-par talks. Some of the talks were about papers that I already read and were published in 2012 or in early 2013 and I do not see the point of listening to a talk about a paper I already read. Some of the talks were about ideas only partially developed, and you could tell it.
The plenary by Trudy Mackay about the genotype-phenotype mapping in Drosophila made me think pretty hard. If in a model (the model) species like Drosophila (with more ~ 200 inbred lines) predictions at the individual level for polygenic traits using SNPs are far from being acceptable (she vigorously supports the idea that epistatic interactions are responsible from the lack of predictive ability from simple additivity), do we have any chance of making right now or in the foreseeable future half-decent predictions for polygenic traits in non-model species? We are very far from it, although Craig Primmer was able to predict correctly 8 out of 10 times the age at sexual maturity in Atlantic salmon using SNPs outliers from a GWAS association study.
Here below, the talks I listened to at the European Society of Evolutionary Biology meeting in Lisbon, 19-24 August 2013. 32 talks total.
Tuesday 20th Aug 2013
De-Meaux: The molecular underpinnings of adaptation in Arabidopsis Thaliana
Ellers: A new perspective in phenotypic plasticity
Foray: Can we predict the effect of thermal fluctuations on specialist and generalist reaction norms?
Morrissey: Selection and evolution of causally-covaryint traits
Duthie: A fluctuating environment drives coexistence in five non-pollinating fig wasps
Reed: Quantifying the demographic cost of selection in a changing environment
Wednesday 21st Aug 2013
Hoglund: Why genetics, genomics, bioniformatics and evolutionary theory may matter for conservation
Geroldinger: The influence of the genetic architecture of a trait on local adaptation in a subdivided population
Chevin: Evolution of discrete phenotypes from continuous reaction norms
Bernatchez: Applied evolutionary genomics in fish conservation: some success stories and challenges for the future
Primmer: A GWAS association study of age at maturity in Atlantic Salmon: implications for conservation and management
De-Cara: Maintaining fitness and diversity in conservation programmes
Otto: Genomic scope of adaptive mutations to different environments
Thursday 22nd Aug 2013
Greaves: Cancer through the eye of evolutionary medicine
Danchin: Genetic and nongenetic inheritance
Friday 23rd August 2013
Mackay: Charting the genotype-phenotype map: lessons from Drosophila
Stearns: Comparing phenotypic, quantitative genetic, and genomic approaches to measuring tradeoffs in a contemporary human population
Hoyle: The fitness implications of adaptation via phenotypic plasticity and maternal effects
Tarka: Evolutionary potential of arrival date on breeding grounds: selection, heritability and microevolution
Whitlock: Equilibrium and non-equilibrium demographic history and the distribution of Fst
Richards: Epigenetic response to novel and changing environments
Fox: Demographic heterogeneity, selection, and population response
Nitsch: Sibship effects on dispersal behaviour
Kvalnes: Estimating fluctuating selection in age-structured populations
Wensink: Indicators of selection pressure and changes in vital rates at multiple ages
Saturday 24th Aug 2013
Lummaa: Sex differences in natural selection on reproductive scheduling and longevity in humans
Coulson: Using evolutionary demography to link evolution and ecology
Chantepie: Actuarial senescence increases the risk of extinction of mammal populations
Lohr: Reduced lifespan and increased ageing driven by genetic drift in small populations
Butlin: Reinforcement
FitzJohn: What drives biological diversification?
Hoekstra: Sexual something (title had symbols, but it was about evolution of sex)
What I am reading
The Sports Gene by Epstein
As a guy who has been involved in sports all his life, I am as certain as a scientist can be (that is, not 100%, but very close) that there are intrinsic differences in sports abilities. It is not only the early environment, not only the social environment, not only motivation or grit or persistence (which, by the way, as behavioral traits should be heritable. That means that there is an "environmental" and a "genetic" component). Looking forward to reading this book.
On getting more efficient - part 1
I thought a lot lately about how to get more efficient at what I do (I hate wasting time and getting side-tracked), and I am going to write a series of posts presenting my workflow, find tons of holes in it as well as finding/testing way to get more efficient.
One of most compelling reasons to go to meetings and workshop is to take some time off from your routine, listening to some folks and get encouraged or discouraged or neutral, try to convince people that your way is the way, and think about what you can do to improve (in a broad sense) your research program, which ends up most of the times to find better and easiest way to publish in top journals and getting cited a lot. No solving world problems or getting a profit here.
I am right now working on 3 model systems (if I remember well, hint: problem), many different topics within the same model systems (including theory), and I expect things to get messier quite soon when I will get the genetic data from marble trout (a month or two, the sequencer just arrived at the lab). I am not teaching and I am quite limited administrative work. Two group meetings a week lasting 1 to 2.5 hrs each. (At least) Part of the weekend (almost always) on.
I am satisfied by my production of ideas and code on a daily basis, including writing a blog post around 5am in a Lisbon hotel because I am jet lagged. I typically work on a one or two model systems a day with a switch that gets quite "formal" most of the time (e.g. I open another workspace in R), and within the same system I typically in a single day on one or two problems. For example, I may work in the same day on kittiwakes (1st model system) and on marble trout (2nd model system). For kittiwakes, I might work on testing measures of growth variability and their effects on survival and reproduction, and for marble trout I might work on estimating the repeatability of body size and on estimating the standard error of parameters on the random effect effect model of growth I am currently working on (2 problems). Plus, for each model system/problem, I typically produce an amazing number of plots, since I am very keen on visual communication to myself and other people. Should I focus on one single problem a day/a week, the same problem for the model system each day/week until solved and ready to get shipped? Or more interesting/fun to work on many problems/model systems at the same time trusting your supernatural abilities of concentration while worrying about the fact that things very clear right now are clear a mud next week (advice to myself: comment code a lot more)?
On Time and Making Figures
How long does it take to make this graph (click to enlarge)?
Yesterday it took one hour and a half. I used the R package ggplot2, which allows for a greater control on all the various plot components, from legend, to points, to lines, than the standard plot libraries (ggplot2 is great and a good fun, by the way). However, ggplot2 is is very verbose and not intuitive (at least for me). Plot quality is very nice, though. The figure above is one of the figures in the soon-to-be-submitted paper on a novel approach to estimate individual variation in biological processes, in this case body growth of fish using the von Bertalanffy growth function (I will explain how the two parameters are positively correlated in a later post). Top row is one fish populations, and bottom row another one.
Anyway, is one hour and a half too much for a single figure? Consider that I am quite skilled with graphs (even if I could not delete the annoying diagonal line in the legends of the density plots - Illustrator will help -, and I still don't know whether is better to use common y-and x-scales for the two populations) and I don't waste much time (I am quite the perfectionist, that's the only problem). I don't know, Hans and I have worked for quite a long time on the process model (in weeks, not in hours), so maybe this is time well spent, even it takes time from doing some more "important" stuff (like actually solving problems, same thing for writing this blog post now that i think about it).
IOF Marie Curie Fellowship – from applying to winning (Part 2)
If you want to have a good shot at the MC Fellowships you need a strong CV, a strong project and credible host insitution/supervisors.
However, it is not enough, and a good “technical” preparation of the proposal goes a long way. When I was checking online for tips coming from previous MC winners, I was lucky enough to land on a webpage (I could not find it later on, probably it was brought down) that was basically reporting the “reviewer’s sheet”, the paper that reviewers had to fill out when judging the proposal. Every question was very specific and it was a repetition of the sub-paragraphs (almost) explicitly requested in the proposal, but in question form. For instance, “has the applicant clearly explained the potential for collaboration between the two institutions?” or “is the potential for the advancement of European scientific research clear?”. I decided to maintain all the sub-paragraph explicitly requested in the proposal and to add in bold the questions (with no bullets or numbers) I found on the “reviewer’s sheet”, but in positive form, that is
Benefit gained from undertaking the project at European Research Area (ERA) level
or
Scientific, technological, socio-economic rationale for carrying out further research in the field
or
Contribution of training provided to diversifying/broadening the competencies of the researcher, and how this will influence the researcher's career development
So, I explicitly stated and explicitly answered.
This helped me prepare a tight and concise proposal where I carefully explained each and every topic of interest for the reviewers. And in case of the repetitions in multiple sections of the same concepts, I simply did not care and carried on.
It was particularly challenging to prepare the “Implementation” section, in which I was meant to describe facilities and structures of the two institutions (UCSC and Polytechnic of Milan), language support, childcare, lodging, transportations, clearly topics I knew very little about. In addition, I never worked at the Polytechnic of Milan, so who knows about transportation and all the rest.
What I did was reading very carefully the websites of the to institutions, where a wealth of information (otherwise skipped) was provided. Language courses, number of apartments/houses available for faculties or research scholars, childcare facilities, everything.
Or how to write the section “Benefit of the mobility to the European Research Area”? Good question. In that case I stressed that Italy, my country of birth and residency and my return country for the reintegration phase of the MC, has been one of the founding members of the European Union and thus “I feel the need to give my contribution to ERA excellence and to Europe to fulfilling its research and innovation potential. I will bring back to the ERA newly acquired skills, knowledge and perspectives and I will transfer them through teaching, great scientific research and communication events and projects.” Maybe slightly off-topic, but quite moving nonetheless.
Last, editing. I do not know how many mistakes I made. Everything all right, do not panic, after reading so many times the same material it is very easy to (subconsciously) skip whole sections and thus missing weird mistakes that are definitely making an impression on reviewers of sloppy preparation. I remember that IPCC (Intergovernmental Panel on Climate Change) became in my proposal the International Project on Climate Change and the Center for Stock Assessment Research (CSTAR) became the Center for Stock Analysis and Recruitment (weird, isn’it?). And I did not even catch them, my colleagues did. So, ask for help, reach out. Offer money, chat time when they break up with the girlfriend/boyfriend, eternal gratitude (does not work that well), dietary advice, share life experiences to create a stronger bond, some insights on the role of Bismarck in the Franco-Prussian war (maybe it works), roof repair, plumbing work, whatever, but ask for help. Ask for 3 hours of reading. And I am not talking about the research part, that should be set well in advance, but just looking for unclear sentences, typos, bad references, formatting, all the stuff that we don’t think it is absolutely essential, but in reality is. Because you can never make a second first impression and sloppiness is not something conducive to a good first impression.
And put in the proposal at least a figure explaining one critical part of the proposal and a GANTT. They help a lot.
PS, I now offer consultations for Marie Curie IF proposals.
Dyson on Oppenheimer
Fantastic Freeman Dyson's review of Robert Oppenheimer: His Life and Mind (A Life Inside the Center) by Ray Monk.
A few excerpts:
"The subtitle, “A Life Inside the Center,” calls attention to a rarer skill in which Oppenheimer excelled. He had a unique ability to put himself at the places and times at which important things were happening."
"He was driven by an irresistible ambition to play a leading part in historic events."
"Einstein never imagined and never accepted this consequence of his theory. Oppenheimer imagined it and accepted it."
- Note: to accept you need to imagine, so Einstein never imagined, full stop, no?
"After learning this, I went to see Oppenheimer and asked him directly why he had thought that tactical nuclear weapons were a good idea. This time, he answered my question. He said, “To understand why I advocated tactical weapons, you would have to see the Air Force war plan that we had then. That was the God-damnedest thing I ever saw. It was a mindless obliteration of cities and populations. Anything, even a major ground war fought with nuclear weapons, was better than that.”
"The real tragedy of Oppenheimer’s life was not the loss of his security clearance but his failure to be a great scientist."
"I believe the main reason why he failed was a lack of Sitzfleisch. Sitzfleisch is a German word with no equivalent in English. The literal translation is “Sitflesh.” It means the ability to sit still and work quietly. He could never sit still long enough to do a difficult calculation. His calculations were always done hastily and often full of mistakes."
"He always wanted to be at the center. This quality is good for soldiers and politicians but bad for original thinkers."
What I am reading
Emotional Equations by Chip Conley.
Thinking Fast and Slow by Daniel Kanheman (re-reading).
Adventures of a Mathematician by Stanislav Ulam (partially re-reading).