Thursday, March 19, 2015

My new preprint is up

As part of their undergraduate training, our students are required to write a short thesis. Usually, due to the paucity of research funding, their theses take the format of a literature review. A few years ago, however, I proposed a computational study to the student I had been assigned. Despite no previous acquaintace with the subject, she eagerly took the task and performed some computations on possible reaction mechanisms of the organomercurial lyase MerB. She only had the time to compute a few of the possible pathways and therefore, after she had written her thesis with the data she had managed to gather, I completed the analysis of  the other pathways we had thought of at the time, and a few that we had not envisaged. Writing it as a paper took me much longer than I had anticipated, mostly because I kept postponing it due to the thrill of running computations on other enzymes and projects. I have now managed to finish it and submitted it to PeerJ, where it is undergoing review. I have made it available as a Preprint, and would be thankful for any comments about it.


Addendum: the paper has been published

Saturday, February 7, 2015

On the wrong use of expressions such as "evolution's null hypothesis"

A new paper published in PNAS has been in the news lately, claiming to have found 2-billion-years old fossils of sulfur-metabolizing bacteria undistinguishable from modern specimens. The abstract is somewhat cautious "The marked similarity of microbial morphology, habitat, and organization of these fossil communities to their modern counterparts documents exceptionally slow (hypobradytelic) change that, if paralleled by their molecular biology, would evidence extreme evolutionary stasis." (emphasis added). In the press release and in their talks with the media, however, the authors of this study have been much more forceful and hyperbolic: they directly claim that these organisms have not changed at all! As any microbiologist worth its salt would attest, it takes a lot more than morphological similarities to establish that two microbial communities are composed of the same species. Otherwise, metabolic tests with dozens of substrates would not be needed to distinguish microbial species: we would simply need to throw the little bugs under a microscope and see what they looked like! How could the authors possibly be sure, simply from their tests, that microbial adaptation to the environment had achieved that of modern bacteria by the time their sample fossilized?

More than this extraordinary leap of logic, I was grated by the author's claim that such a lack of evolution would be in agreement with evolution's null-hypothesis of no biological change in the absence of changes in the physico-chemical environment, and it therefore strengthens the case for evolution.... How is it possible to cram so many errors and inaccuracies in such few words? How could the peer-reviewers let such inane nonsense appear in the title of the paper? Let us start to unravel the many mistakes in this formulation:

  • What the authors call "evolution's null hypothesis" has NOT (as far as  I have been able to ascertain) ever been claimed as "evolution's null hypothesis" at all: it is well-know, at least since the seminal work by Kimura, that the strongest driver of genetic variation is not the positive selection of advantageous mutations but the random fixation of neutral (or barely neutral mutations). Indeed, in humans only ca. 400 of the estimated 16500 genes show strong evidence of positive selection, even though all of the genes show variation from those of closely-related species.  It is therefore NOT at all expected that genomic stasis would be observed over a long period of time. Stating (as the authors) that observing no  change in these organisms is a confirmation of the mechanisms of evolution reveals a shocking lack of knowledge regarding  molecular evolution. And the authors have not even proved that there was no change: that would require establishing that their ATP-producing metabolism is as efficient as that of their modern counterparts, that they are able to use the same substrates, and contain all the same  enzymes, etc....
  •  By claiming that an unchanging environment leads to an immutable species, the authors commit a further logical fallacy: after all, there was a time (let's call it t0) when the ancestor to that community first entered that unchanging environment. If an unchanging environment leads to evolutionary stasis, then the authors are claiming that at time t0+1 million years the species would be equal to that at time t0, or that at t0+2 million years, and so forth. But of course adaptation to an environment is not instantaneous, unless the parent ancestor already possesses all enzymes needed to thrive there (and this is most unlikely, as there has been no selective pressure for that). An unchanging environment therefore causes evolutionary pressures, at least in regards to the first cells which venture there.
  • When an observation is compatible with different theories, it cannot be used to further any of them: after all, seeing no change in 2 billion years could also be used to argue for the immutability of species. It is therefore logically fallacious to present it as proof that Darwin was right. Also, evolutionary theory has also developed a lot since the writing of "On the Origin of Species by Means of Natural Selection, or the Preservation of Favoured Races in the Struggle for Life". Shouldn't other workers, like Kimura, Felsenstein, Farris and Gould be acknowledged? 
  • Why would any scientist need to claim that “the findings therefore provide further scientific proof for Darwin’s work.”? Do astrophysicists need to state “the findings therefore provide further scientific proof for heliocentrism” every time that a new comet is found, its orbit is computed, and it is found to move around the sun rather than around the earth? Do anthropologists working in the Balkans need to point out that “the findings therefore provide further scientific proof that human societies do not all resemble hunter-gatherer groups?” The curious insistence of American-based media to frame biological discoveries as a supposed debate/beauty-contest between "evolution" and "creationism/immutability of species/intelligent design" is completely mind-boggling to any European, whether religious or not. This insistence was also displayed in Neil de Grasse Tyson's "Cosmos", which, unlike Sagan's masterpiece, seemed more interested in scoring debate points against a sub-section of its domestic audience than on presenting the astounding amount of knowledge mankind has gathered in the few millenia we have spent since the dawn of agriculture.



Claims unwarranted by data, exaggeration and PR stunts: all of these are usually as ascribed (rightly or not) to politicians, polemicists, salespeople and shady companies seeking to attract capital. Do we really want science to be tarred by the same brush?

Monday, September 8, 2014

Making good on my "Open Access" pledge


My most recent paper has just been published in PeerJ . It was a LONG time in the making, to the point that my 12-yo daughter once told me (only half-in-jest), that I should "cut my losses and forget about it". I am quite happy about how it turned out: besides describing an analysis of a reaction mechanism and the influence of the redox state of a hard-to-converge Fe-S cluster , it also contains  the first computations including the weighed contributions of 1.2*1013 protonations states of a protein on the reaction it catalyzes. The computational approach described here is relatively simple to perform provided that one has a good estimate of the relative abundances of those protonation states, which can be obtained through Monte Carlo sampling  once the site-site interactions have been computed with a Poisson-Boltzmann solver. To my mind, this is clearly superior to the usual approach of considering only  the "most likely" protonation state (which may often not be the state with the most significant influence on the electrostatic field surrounding the active site). What do you think of it?


Programs needed to use this approach:
MCRP, by Baptista et al., ITQB, Lisbon
MEAD, by Don Bashford, currently at St. Jude Children's research hospital
Any molecular mechanics code, to compute the change of the total electrostatic energy as each individual amino acid is protonated/deprotonated



Tuesday, July 22, 2014

Challenges of teaching Biochemistry to Health Sciences students

Had anyone told me, 20 years ago, that I would earn my living as a lecturer, I would have considered it as a put-down. I did have a lot of respect and appreciation for (most of) my lecturers at the University of Porto, but I expected to become a full-time scientist, rather than a "lecturer who finds time to do some science in-between classes/grading" or a "researcher with required part-time lecturing duties". Real life disabused me of that expectation: due to the dearth of other scientific jobs in Portugal, I did become a "lecturer who finds time to do some science in-between classes/grading" after finishing my PhD. 

The culture shock I experienced when I first lectured to Health Science Students left me unable to speak about much more than the woes  of teaching for the best part of a year. A big portion of my surprise came from my first-contact with regular students who attended my lectures simply because they were required to by the University, rather than due to recognizing the subject as a relevant background for their (mostly) vocational training as Physical Therapists, etc. Being required to take most classes to graduate  (rather than choosing large part of the curriculum around a core subset) is a very common feature of university curricula in Portugal. In principle, it is meant to ensure that all students have a balanced curriculum and do not "flee" the hardest subjects. In practice, it also tends to lead to ever larger classes of those same "hard subjects", since students tend to consider those lectures as bureaucratic hurdles thrown at them, rather than as valuable knowledge and therefore feel disengaged, alienated and fail them in large numbers.

All classes I have taught (Biochemisty, Organic/General/Analytic Chemistry, Basic Mathematics/Statistics) fall into the "hard subjects" class. In my first years of lecturing, I had a most demotivated cohort of students. My expectations regarding their performance were generally very unrealistic, as my baseline comparison was my own student experience at my "alma mater", where I was surrounded by engaged student peers who were motivated into learning pure scientific subjects, and did not regard them as "filler" or "bureacratic hurdles" aimed at winnowing the sutdent body. Moreover, my "alma mater", the Faculty of Sciences at the University of Porto, was famous among students by its harsh grades: attrition was relatively high, less than 20% of those graduating from its Chemistry or Biochemistry curricula would have an average grade of 16/20 or 17/20, and higher final grades were virtually unheard of. In other Portuguese Universities, final average grades of 18/20 were common, even though their student body was of the same (or even slighly lower) quality, as judged by their entry grades.

I eventually adapted to the students' expectations, and developed a teaching method that engages students and apparently motivates them (as judged from the appreciative comments in teacher evaluation forms). However, I find that this only seems to work during class time: students pay attention, seem to be making all the right connections (as long as I softly nudge them towards the right path, etc.), congratulate me on the quality of lectures, etc. In tests/quizzes/exams, however, a strong disconnect appears: ca. 50% of my students still struggle with many concepts that I would consider as absolutely basic. Why does this happen?

I have just found out that there is a proper name to what is happening in my classes: pseudoteaching (defined as " The concept [...] that even the most outwardly perfect lesson can result in students not actually getting what it is you wanted them to understand."): along with this, there is also pseudolearning ("Going through the “expected” steps without extracting a solid, working understanding of a topic would") and pseudostudying (which I would define as "reading and working the material to the point where one feels tired  but without actually taking anything from the exercise due to inability to distill the core concepts into working knowledge").  I cannot prevent students from pseudolearning or pseudostudying (apart from exhorting them to rest properly, keep their blood sugar levels up while studying and work/study in short bursts daily rather than pulling all-nighters on the eve of the tests). Avoiding pseudoteaching is in my power, but I do not (yet) know how to: Jan Jensen (following Mazur) advocates a flipped classroom model where exposition occurs outside class time (using short video lectures and key exercises) followed by solving exercises "in-class" with free exchange of ideas among students (peer-instruction). I do not think this  method would help with my students, though: a previous experience of short (< 10 minutes) in-class quizzes led to class disruption, acutely stressed students during and after the quiz and minimal improvement in weekly off-class engagement with the study material. What would you suggest me do?

Thursday, July 17, 2014

Do you want to publish for free in PeerJ?

I started following the Open Access movement ca. 2 yrs ago, mostly through the blogs of Michael Eisen, Jan Jensen and Mike Taylor. I was obvioulsy well aware of the successful OA outfits, like PLOS and BiomedCentral, but had never considered publishing there due to the shortage of funds and the non-reimbursability of such expenses by my country's Science Foundation. I joined PeerJ shortly after they "opened for business", due to their their very small fees and because of commitment to transparent peer-review , which to my eyes sets them apart from the wide number of OA venues which spam email boxes daily all over the world. 
After publishing my first paper on PeerJ, I have received five referral codes from them, each of which entitles an author to the free publication of a paper in PeerJ. The codes expire on August 10th. Should you wish to take advantage of one of these, please drop me a line. PeerJ only accepts submissions on the area of Biology, which is defined very broadly (from ecology and paleontology to virology, bioinformatics and computational biochemistry).

Addendum (July, 21st, 2014): Four codes have been distributed. One left to go

Addendum (September, 14th, 2014): I have received five new codes, valid through October 2nd, 2014. Any takers?

Tuesday, June 17, 2014

Femex 2014 in Oslo has just ended

I thoroughly enjoyed the second "Promoting Female Excellence in Theoretical and Computational Chemistry" conference, both in choice of speakers and convivial atmosphere. The ratio of female/male speakers obviously favored females, but considering the nature of the conference I would not mind if it were even more skewed towards the female end of the scale, as it probably encourages audience questions from women. The format was very adequate, but ratio between established/early-career speakers might probably be adjusted slightly (for example by including a dozen more  presentations selected from poster submissions) to enable increased "name-recognition" of younger researchers.



An extra day or two, and some more free time for socialization would have been very welcome: I found that the enthusiasm and conversation flow increased substantially after the banquet talk, but by that time the meeting was coming to an end and productive conversations had to be cut short due to the need to catch the flights home. If the "after-banquet talk" could be moved to the first night of the conference, the focus of conversations during the meeting might have included more reflexion on the sociology of our profession, the way that the subtle biases which discourage hiring scientists with a publication-gap of a few years are built/accepted/torn down, and so forth. That talk did serve as a wonderful conversation starter.



I loved the presence of  children in the meeting, and think that a specific sentence in the conference website stating that they are welcome to the conferences would have a positive effect in lowering barriers to attendance, and in removing the prevalent "productivity-minded" biases which make graduate students, post-docs, non-tenured faculty feel that embracing a scientific career must lead to a neglect of other important parts of life. No matter how many "empowering" talks, positive discrimination, awareness campaings, etc., an academic culture where powerful figures of authority (whether star professors, PIs or funding agencies) demand or expect that researchers put their personal life behind their scientific productivity  skews the resulting researcher pool towards the obsessively-driven, hyper-ambitious, un-empathic tail of the population spectrum.  Whether that tail is mostly male, mostly female, or "equal-opportunity", it favors non-collegial behavior and chases good people away. Hyper-ambitious researchers may be very productive, but they cannot produce much science if their behavior leads to talented people fleeing towards other endeavors.

Congratulations to the organizing team, and a heartfelt "thank you" to all participants.

Thursday, April 24, 2014

Gamess (US) frequently asked questions part 6: Obtaining proper SCF convergence (Anti-)ferromagnetic coupled Fe-S clusters

Obtaining SCF convergence of FeS clusters is a very demanding task.
The problem in FeS clusters is the arrangement of spins on the Fe atoms: if you have a cluster with 4 Fe atoms, each of them with 5 up-spins, and a total spin of zero, the arrangement of spins on the atoms could be
  • Fe1 and Fe2  up-spin, Fe3 and Fe4 down-spin; or
  • Fe1 and Fe4  up-spin, Fe2 and Fe3 down-spin; or
  • Fe1 and Fe3  up-spin, Fe2 and Fe4 down-spin;
The problem is compounded if you have a mixture of Fe2+ and Fe3+, which may lead to 12 (or more) different spin arrangements, depending on the number of Fe2+ atoms. However, if you have a good guess SCF for one instance instance, you may simply substitute the coordinates of Fe2 with those of Fe4 to get a comparably good guess for the second instance, and so forth... This is the approach suggested by Greco, Fantucci, Ryde, de Gioia (2011) Int. J. Quantum Chem. 111, 3949-3960. Obtaining the guess for one of the instances is in itself quite difficult, and I usually follow the approach outlined by Szilagyi, R. K. and Winslow, M. A. (2006) J. Comput. Chem., 27: 1385–1397  .
It goes like this:

- obtain orbitals for bare Fe2+, Fe3+, S2-, and isolated ligands, with proper spins on the Fe atoms (5/2 for Fe3+, 2 for Fe2+)

- Manually split the "alpha/up" and "beta/down" portions of the resulting  $VEC groups. For example, assuming you have a system with three Fe atoms (two Fe2+ and one Fe3+) with total spin S=5/2 and the $VEC groups for bare Fe2+ and bare Fe3+, you should cut the $VEC groups of Fe2+ and Fe3+ as:


$VEC  for the alpha (up) electrons of Fe2+   (let's call it "Fe2+_5_d_electrons")
$VEC  for the alpha (up) electrons of Fe3+   (let's call it "Fe3+_5_d_electrons")
$VEC  for the beta (down) electrons of Fe2+   (let's call it "Fe2+_1_d_electron")
$VEC  for the beta (down) electrons of Fe3+   (let's call it "Fe3+_0_d_electrons")
The total spin S=5/2 in this sample problem implies that  both Fe2+ atoms spins should annull each other, i.e., one Fe2+ is mostly "up" and the other is mostly "down". Building the new guess for the "up" electrons should therefore include:

"Fe2+_5_d_electrons" for one of the  Fe2+ ions,
"Fe2+_1_d_electrons" for the other  Fe2+,
"Fe3+_5_d_electrons" for the Fe3+

Building the new guess for the "down" electrons should  include:
"Fe2+_1_d_electrons" for the FIRST Fe2+ ions,
"Fe2+_5_d_electrons" for the other Fe2+,
"Fe3+_0_d_electrons" for the Fe3+


- combine the orbitals using the small utility called combo, which you may obtain from Alex Granovsky's Firefly website.

- Manually paste the "alpha" and "beta" guesses  into a single $vec group, which would be the proper guess.

- cross all your fingers and toes, and expect it to converge into the proper state. If it does not converge, change convergers (SOSCF=.T. DIIS=.F.), onset of SOSCF (SOGTOL=1e-3) , etc.

- after SCF optimization using this guess, manually scramble the ordering of Fe atoms in your input, to ascertain whether a lower energy solution can be obtained with a different spin distribution.



Good Luck!