The Good: lightning talks give more speakers a chance to present their material to an engaged audience; they cultivate succinct speaking skills. If you don’t like a talk in the session, you don’t have to wait for half an hour for the next one, you only have to wait for five minutes.

The Bad: a long session crammed with lightning talks may cause a jumble in the typical audience member’s brain. Talks that are early or late in the session may  receive more attention due to the serial position effect, so that the middle talks are completely lost in the muddle, and the first and last couple of talks are those that are remembered.

Still, suppose you submitted an abstract to a conference, and made the cut for a lightning talk; what now?

Forget most of the skills you were taught for a regular 20 minute conference presentation, or 40 minute seminar. Lightning is a different beast. In a long talk, you teach a bit (background to the field & introduction to the problem at hand), show your stuff (your work), and advertise (show how your work contributed to the field, and how you left it better).

In a lightning talk, you want to get a single message across. And you want it to stand out. So you cannot afford to be too complex, you just don’t have the time.

Do: Prepare five to ten slides. Make sure they are sparse. An image or two per slide. No complex graphs. If you need words, write them big and few.

Do introduce yourself clearly at the beginning  (name, affiliation, position, what you do)

Do  clearly introduce whatever you are presenting.

Do give the acknowledgement slide at the beginning   Although that is common practice in regular talk to give it at the end, in a lightning talk you want your last slide to be something else. See below.

Do speak at your normal pace.

Do make the last slide the impressive one: clear, strong message that will linger a minute longer during Q&A time, impressing itself upon the audience before it is time to move to the next talk. You do not want the acknowledgement slide to be last, as is traditionally done in longer talks.

Do: rehearse, rehearse, rehearse. Even if you are an accomplished speaker who can do a long talk without rehearsals, the lightning talk is a different beast. Waffling costs precious seconds, Moreover, getting back on track you may be tempted to speak faster to make up for lost time. Which is a no.

Don’t cram too many slides or be tempted to speak too fast. Find a way to convey your message at a normal speaking pace. Compressing more words into less time does not increase the information you convey, it actually deceases it. People can only process so much at a given time. Remember that a talk, including a lightning talk is about making people understand something new, not about you  maximizing words-per-minute.

Don’t go over the allotted time.  If you are not finished by the time the clock buzzes or the session chair signals you to get off, just say “sorry, time’s up. Catch me at the coffee break if you want to hear more.” — and step off the podium .

Ride the lightning!

Short quiz and a movie for DNA day.

1) We celebrate DNA day because:

a) Congress said so

b) Francis Collins said so

c) I said so

2) Who has DNA?

a) CSI Miami

b) James Watson

c) Please, please, PLEASE let the  paternity test comes back negative…

3)  Nature vs. Nurture: which is more important?

a) Nature

b) Nurture

c) Nurture, but only if your mother was a hamster and your father smelled of elderberries

4) The following movie shows:

a) Replication

b) Application

c) Fumigation

Keynote speakers:

• Patricia Babbitt, University of California, San Francisco. Protein similarity networks: Identification of functional trends from the context of sequence similarity
• Alex Bateman, European Bioinformatics Institute Using protein domains and families for functional prediction
• Anna Tramontano, “La Sapienza” University, Rome. TBA

Key dates:

• April 20, 2013: Deadline for submitting extended abstracts posters & talks
• May 9, 2013: Notifications for accepted abstracts e-mailed to corresponding authors
• May 16, 2013: Deadline for presenters to confirm acceptance of invitation to speak.
• July 20, 2013: AFP SIG preceding ISMB/ECCB 2013, Berlin.

Sequence and structure genomics have generated a wealth of data, but extracting meaningful information from genomic information is becoming an increasingly difficult challenge. Both the number and the diversity of discovered sequences are increasing, and the fraction of genes whose function is known is decreasing. In addition, there is a need for annotation which is standardized so that it could be incorporated into function annotation on a large scale. Finally, there is a need to assess the quality of the available function predictionsoftware.

For these reasons and many more, automated protein function prediction is rapidly gaining interest among computational biologists in academia and industry.

The Automated Function Prediction Special Interest Group (AFP SIG) has been part of ISMB since 2005. We call upon all researchers involved in gene and protein functionprediction and annotation, both computational and experimental, to submit an abstract to the AFP meeting. Authors of select abstracts will be invited to give a talk and/or present a poster.

We will also be discussing the upcoming second Critical Assessment of Function Annotations, or CAFA 2. CAFA 1 was a highly successful experiment, engaging 30 groups worldwide, and has resulted in 16 peer-reviewed papers in Nature Methods and BMC Bioinformatics.

We are looking forward to a new and expanded CAFA 2 in 2013-2014, which will include a cellular component prediction track, and a human-specific track.

For further instructions on AFP 2013, please go here: http://BioFunctionPrediction.org

Please submit your abstract now, we are looking forward to seeing you in Berlin.

For continuing information, please subscribe to the following Google Group: https://groups.google.com/forum/?fromgroups#!forum/afp-cafa

Contact: afp.cafa.2013@gmail.com

Organizers:

• Iddo Friedberg, Miami University, Oxford, OH USA
• Sean Mooney, Buck Institute for Aging Research, CA USA
• Predrag Radivojac, Indian University, Bloomington IN, USA

Steering committee:

• Steven Brenner, University of California, Berkeley, USA
• Patricia Babbitt, University of California, San Francisco, USA
• Christine Orengo, University College London, UK
• Burkhard Rost, Technical University Munich, Germany

Program committee:

• Mark Wass, Kent University, UK (chair)
• Iddo Friedberg, Miami University, OH, USA (co-chair)

But WSJ editorial from E.O. Wilson has irked me so much, I have decided to go for it. The upset I felt when reading this was on several levels: as a teacher, and a scientist, and as a person concerned for the future of science, and science literacy.  In this editorial, Wilson promotes a type of scientific illiteracy that is dangerous if taken to heart by aspiring scientists.

In essence, Wilson draws from his personal experience as a successful scientist who is not only semi-illiterate in math, but proud of it. He claims that, if he succeeded as a math illiterate, so can other scientists, except in ”a few disciplines, such as particle physics, astrophysics and information theory.”   (All quotes are from said article, unless noted otherwise.) He claims that “Far more important throughout the rest of science is the ability to form concepts, during which the researcher conjures images and processes by intuition.”  He continues to state that: “ The annals of theoretical biology are clogged with mathematical models that either can be safely ignored or, when tested, fail. Possibly no more than 10% have any lasting value.”

OK, let’s take a closer look. First, Wilson’s claim from personal experience that “advanced math” is unnecessary. That he only got to taking calculus when he was tenured faculty at Harvard at the age if 32, and was never more than a C student. He became successful anyway. Wilson is a successful outlier: a genius that did the right things and was in the right place and time to get where he got, the lack of some basic skills otherwise mostly needed not hindering him.

This is the same as other successful extreme outliers: Bill Gates who became a billionaire but who never completed college. Gates is held as a shining example for those who argue that you don’t need college, all you need is hard work and ingenuity. (If everyone was ingenious, then it won’t be a rare and advantageous trait anymore, would it? And if everyone as a billionaire, that would mean we are in the midst of a Weimar Republic hyper-inflation.) This rationale has led to thousands of engineering school drop-outs to form their own startups in the late 1990′s, and we know how that has ended. For every one Bill Gates, there are thousands of college dropouts that went broke, and were unemployable due to a lack of formal skills. On the other hands, there are the thousands who completed their degree, gained successful work experience, and some may have formed their own startups that may or may not have become successful, but who have a formal and comprehensive education to draw upon and make them employable.

The same principle applies in Wilson’s case: for every science professor that is a self-professed math illiterate and who made tenure in Harvard at the age of 32, there are thousands of postdocs who are vying for an assistant professorship in any university. And those postdocs need to know some statistics if they are to be able to design and interpret any kind of basic experiment. This is a basic tenet of any experimental science  and any advice to the contrary is terrible.  They need to know calculus if those experiments have a time course. And they need to know basic programming if they are to analyze large amounts of data. Universities can afford to be selective.  The postdoc who will usually get the job will be the one offering the most promise. And the most promise is offered by displaying a broad range of skills (among other things, of course). Furthermore, even basic skills are required to graduate with a viable Ph.D. It is a poor degree in biology whose owner does now know how to test for statistical significance, that most basic of requirements for designing and conducting experiments. Furthermore, biology today has changed in many ways since Wilson was tenured in Harvard. It is mostly a data rich science. “Intuition” cannot be used when gigabytes of data are involved.

The other problem I have with Wilson’s take on math, is his categorization of math as an auxiliary discipline which only serves to formalize creative ideas, but which is not part of the creative process.  ”Ideas.. emerge most readily when some part of the world is studied for its own sake. They follow from thorough, well-organized knowledge of all that is known or can be imagined of real entities and processes within that fragment of existence.” Wilson claims that knowledge of math by itself cannot help generate such ideas, but can only help to formalize them once the nebulous intuition generates them.  He draws upon Darwin as the ultimate biologist with no math background who “made it”. Darwin is Wilson’s Bill Gates, his outlier from which he draws general conclusions. However, even Darwin himself wrote:

I have deeply regretted that I did not proceed far enough at least to understand something of the great leading principles of mathematics;  for men thus endowed seem to have an extra sense.

It is this extra sense that makes math such a powerful asset to biological thinking. Indeed “thorough, well-organized knowledge of all that is known or can be imagined of real entities and processes within that fragment of existence.” requires math to chart and define these “processes” as processes in the first place. That “well-organized knowledge” can become well-organized by better understanding the fundamental discipline of knowledge organization: statistics. In fact, Wilson uses an intuition of math, unwittingly,  to have these “processes” emerge. He then finds someone for the “follow-up steps” which “usually require mathematical and statistical methods to move the analysis forward.” Imagine what would be possible if Wilson was possessed of that “extra sense” that Darwin has recognized as promoting the creative process, rather than treating math as  a technical followup. The ability to come up with models not relying on “dreaming” alone, but drawing upon the “extra sense” that Wilson unfortunately and mistakenly spurns as a mere adjunct discipline. Actually, there is no need to imagine such a scenario. As there are so many discoveries made in biology that drew upon both math and empirical biological knowledge. Ecology, genetics, population biology, neurobiology, biochemistry and, of course, molecular biology.

There is, of course, the remote possibility that Wilson is trolling for the sake of promoting quantitative thinking which is lacking in too many corners in science and especially in biology. That Wilson has recognized the lack of math in certain disciplines is hurting and holding back those disciplines, and has decided that the best way to promote young biologists to learn math and adopt it to their research and in their teaching is by provoking a tribal war between the mathematically literate the those who are not. Sadly, I doubt that this is the case. More likely, his self-professed math illiteracy serves the purpose of not recognizing the generalization from an outlier cannot serve as a viable model, or even an argument to support his position.

Finally, for a far more comprehensive post than mine, which links to many other responses including those supportive of Wilson, I suggest you read E.O. Wilson vs. Math at Dynamic Ecology.

It seems like the forces of light have triumphed somewhere around September 2006:

…as have their evil counterparts, April 2009:

bacteria are neck-in-neck with humans:

But they beat the largest creatures on Earth:

Of course, you can’t beat cats:

mydict = {'a':5, 'b':2, 'c':1, 'd':6}

You want to sort the keys by the values,  maintaining the keys first in a list of tuples, so that the final list will be:

[('c',1), ('b',2), ('a',5), ('d',6)]

aaaand, the stupid Python trick involves a nested list comprehension:

sorted_list = [(k,v) for v,k in sorted(
                 [(v,k) for k,v in mydict.items()]
                 )
              ]

To get a reverse sorted list:

[('d',6), ('a',5),('b',2),('c',1)]

[(k,v) for v,k in sorted(
   [(v,k) for k,v in mydict.items()],reverse=True
   )
]

Crikey. That’s a stupid python if I ever held one!

Thanks to Mitch Balish for calling my attention to this one.

SR1 bacteria are not exactly a household name, even among microbiologists. They were first discovered in contaminated aquifers,  and since then they were found to be also in animal and insect guts, as well as in human mouths. They are even suspected of being a cause of periodontal disease.  I should probably say here that SR1 is a whole phylum of bacteria, and not a single genus or species. The reason that they are not that well known is that their discovery was fairly recent.

Also, no one has ever actually seen or grown SR1.

All we know is that they are called SR1

Like 99% of the microbial world, SR1 cannot be cultured in the lab. Maybe they need the close association of other bacteria that provide nutrients or some sort of essential function that SR1 lack. (See the “Black Queen Hypothesis” on how members of multi-species bacterial communities provide essential functions to the community.)  Our knowledge of SR1 comes from finding molecular species markers in environmental genomic studies. These markers give us an estimate of the prevalence of SR1 bacteria. The fact that we find them in certain types of environments tell us of their preferences for low oxygen. Some also are suspected to be free living, while those that are associated with humans and other animals are thought to be parasitic or commensal.

Characterizing bacteria by small fragments of genomic sequences is like walking in a forest and not seeing any animals. You only see strange tracks on the ground, and  branches with clumps of fur that were scraped off several animals. Judging by the fur and tracks, you can tell that these may be animals you haven’t seen before. Even though you cannot observe the animals. you can lean some things about them: fur color an consistency can help you classify them in a certain mammalian family (only mammals have fur). The tracks and height of fur clumps can tell you the animals’  approximate size and weight. If the tracks are clawed, you may infer the mystery animals are carnivorous. If they are hooves, they are probably herbivorous.  You can also infer their range, and perhaps an approximate number of individuals, and whether different subspecies likes to live in the hilly or the flat part of the forest. An occasional feather on the ground, or nest in the trees may tell you there are birds in th forest too. Still, you can’t learn much more than that there is a community of animals, and the approximate number of species and some basic traits they may have.

When you see it…

So a new study on the single-cell sequencing of SR1 is quite welcome: it’s a good trap for our unidentified animals, or removing the helmet from the Stig’s head.  Metagenomics can provide a sample-based genetic picture of a microbial community, but can’t tell us of the structure of an individual bacterial genome.  A single whole genome can inform us of what a bacterium can or can’t do, as an individual biological unit. So we can learn, for example, why members of SR1 cannot be cultured, or why they may be involved in periodontal disease.

A group from Oak Ridge National Laboratory, Yale University and the Joint Genome Institute have  managed to do that. They isolated single cells of  SR1 bacteria, and although the bacteria could not be cultured, their genomes could be sequenced. Once the researchers actually looked at the predicted genes, they found that they were uncharacteristically  and unrealistically short.  The UGA codon that normally signals the stop of protein translation, appeared with an alarmingly high frequency, cutting genes in the middle. This led them to think that the UGA codon may not actually be a stop codon. Searching the genome, they found a gene for tRNA that looks like a Glycine carrier, and had the ACU anticodon.

Bingo. The UGA codons in SR1 actually code for Glycine.  An (almost) first. UGA for glycine were found before in mitichondria of a sea-squirt, but only there.

The genome the researchers sequenced is incomplete, but there are several other interesting things they found. One is no evidence of respiration-related genes, explaining why SR1 are found in oxygen-poor environments.  Another oddity is a strange kind of  RubisCO. RuBisCO is probably the most abundant protein on earth, and is the enzyme involved in the first major step of converting CO2 into sugars – something plants and phtosynthetic bacteria do. SR1 are not photosynthetic, and their RubisCO, although able to fix carbon, is actually involved in the breakdown of sugars. This  ”re-purposing” of RubisCO is something that is found in archaea, but rarely in bacteria. Another first.

This study is exciting, because it shows the power of single-cell genomics. No matter how much metagenomics you do, you will still be left with an incomplete picture of the genomic data in your sample, and left guessing as to what individual species may be doing. With single-cell genomics, “complete picture” genomics is back, and it’s taking on the uncultured microbes.

Campbell, J., O’Donoghue, P., Campbell, A., Schwientek, P., Sczyrba, A., Woyke, T., Soll, D., & Podar, M. (2013). UGA is an additional glycine codon in uncultured SR1 bacteria from the human microbiota Proceedings of the National Academy of Sciences DOI: 10.1073/pnas.1303090110

A new journal, Molecular Metabolism has the following policies: one week for reviews, and three possible outcomes only: Reject, Accept, or Minor Revision. Good for them on both decisions. Bonus: your editors are  Mr. Blonde, Mr. Blue, Mr. Brown, Mr. Orange and Mr. Pink. And they are professionals (although they may not tip).

Keynote speakers

Celebromics: sequencing the genome or microbiome of a celebrity. Generally the sequence is not even published, but just the act of sequencing it provides publicity for the lucky lab, the celeb, and maybe even a microbial species or two. “They sequenced the genome of Keith Richards, and found a duplicated set of multiple drug resistance genes.”

Contaminomics: sequencing results published prematurely, and later discovered that the major finding is the result of a contamination.

DuhOmics: unsurprising results from a genomic study. Usually confirming common knowledge that did not require a genomic study in the first place. 

Lazarusomics: sequencing the genome of an extinct animal, including hominids, with the implicit or explicit promise that we will be able, very soon, to reverse the extinction.

Shockomics:  related to advertisomics. Sequencing for shock value and pop publicity. Usually involving human or animal bodily secretions or parts you’d rather not have known about.

TooMuchInformationOmics: A result of the personal genomics and microbiome industry. No, I am not interested in that heel spur gene that you got from your grandmother, nor am I interested  in the novelty of the chlamydia strain they found in your partner’s microbiome.

ZZZomics: an omics paper that makes you fall asleep half way through the introduction.