Stephen Hawking's advice for twenty-first century grads: Embrace complexity

Charles Joseph Minard's famous graph showing the decreasing size of Napoleon's Grande Armée as it marches to Moscow; a classic in data visualization (Image: Wikipedia Commons)

As the economy continues to chart its own tortuous, uncertain course, there seems to have been a fair amount of much-needed discussion on the kinds of skills new grads should possess. These skills of course have to be driven by market demand. As chemist George Whitesides asks for instance, what's the point of getting a degree in organic synthesis in the United States if most organic synthesis jobs are in China?

Upcoming grads should indeed focus on what sells. But from a bigger standpoint, especially in the sciences, new skill sets are also inevitably driven by the course that science is taking at that point. The correlation is not perfect (since market forces still often trump science) but a few examples make this science-driven demand clear. For instance if you were growing up in the immediate post-WW2 era, getting a degree in physics would have helped. Because of its prestige and glut of government funding, physics was in the middle of one of its most exciting periods. New particles were literally streaming out of woodwork, giant particle accelerators were humming and federal and industrial labs were enthusiastically hiring. If you were graduating in the last twenty years or so, getting a degree in biology would have been useful because the golden age of biology was just entering its most productive years. Similarly, organic chemists enjoyed a remarkably fertile period in the pharmaceutical industry from the 50s through the 80s because new drugs were flowing out of drug companies at a rapid pace and scientists like R. B. Woodward were taking the discipline to new heights.

Demand for new grads is clearly driven by the market, but it also depends on the prevalence of certain scientific disciplines at specific time points. This in turn dictates the skills you should have; a physics-heavy market would need skills in mathematics and electronics for instance, a biology-heavy market would mop up people who can run Western blots and PCR. Based on this trend, what kind of skills and knowledge would best serve graduates in the twenty-first century?

To me the answer partly comes from an unlikely source: Stephen Hawking. A few years ago, Hawking was asked what he thought of the common opinion that the twentieth century was that of biology and the twenty-first century would be that of physics. Hawking replied that in his opinion the twenty-first century would be the "century of complexity". That remark probably holds more useful advice for contemporary students than they realize since it points to at least two skills which are going to be essential for new college grads in the age of complexity: statistics and data visualization.

Let's start with the need for statistics. Many of the most important fields of twenty-first century research including neuroscience, synthetic and systems biology, materials science and energy are inherently composed of multilevel phenomena that proliferate across different levels of complexity. While the reductionist zeitgeist of the twentieth century yielded great dividends, we are now seeing a movement away from strict reductionism toward emergent phenomena. While the word "emergence" is often thrown around as a fashionable place-card, the fact is that complex, emergent phenomena do need a different kind of skill set.

The hallmark of complexity is a glut of data. These days you often hear talk of the analysis of 'Big Data' as an independent field and you hear about the advent of 'data scientists'. Big Data now has started making routine appearances in the pharmaceutical and biotech industry, whether in the form of extensive multidimensional structure-activity relationship (SAR) datasets or as bushels of genomic sequence information. It's also important in any number of diverse fields ranging from voter behavior to homeland security. Statistical analysis is undoubtedly going to be key to analyzing this data. In my own field of molecular modeling, statistical analysis is now considered routine in the analysis of virtual screening hits although it's not as widely used as it should.

Statistics was of course always a useful science but now it's going to be paramount; positions explicitly looking for 'data scientists' for instance specifically ask for a mix of programming skills and statistics. Sadly many formal college requirements still don't include statistics and most scientists, if they do it at all, learn statistics on the job. For thriving in the new age of complexity this scenario has to change. Statistics must now become a mandatory part of science majors. A modest step in this direction is the publication of user-friendly, popular books on statistics like Charles Wheelan's "Naked Statistics" or Nate Silver's "The Signal and the Noise" which have been quickly devoured by science-savvy readers. Some of these are good enough to be prescribed in college courses for statistics non-majors.

Along with statistics, the other important skill for students of complexity is going to be data visualization and formal college courses should also reflect this increasingly important skill set. Complex systems often yield data that's spread over different levels of hierarchy and even different fields. It's quite a challenge to visualize this data well. One resource that's often recommended for data visualization is Edward Tufte's pioneering series of books. Tufte shows us how to present complex data often convoluted by the constrains of Excel spreadsheets. Pioneering developments in human-computer interaction and graphics will nonetheless ease visual access to complicated datasets. Sound data visualization is important not just to simply understand a multilayered system or problem but also to communicate that understanding to non-specialists. The age of complexity will inherently involve researchers from different disciplines working together. And while we are at it it's also important to stress - especially to college grads - the value of being able to harmoniously co-exist with other professionals.

Hawking's century of complexity will call upon all the tools of twentieth century problem solving along with a few more. Statistics and data visualization are going to be at the forefront of the data-driven revolution in complex systems. It's time that college requirements reflected these important paradigms.

First published on the Scientific American Blog Network.

Some thoughts on the events around Boston

We were enjoying a quiet evening of music and reading on Thursday when my wife alerted me to a message she got from the MIT emergency system that there had been a shooting somewhere on the campus. A while later we came to know that a police officer had been shot and killed right in front of my wife's department. After making sure that folks we knew from MIT were safe, we stayed awake for about two more hours reading the news updates. By the time we went to sleep we had found out that there was a connection between the shooting of the MIT police officer - a promising young man who later tragically died - and the Boston marathon bombing.

When we woke up the next day the situation was a little surreal: "Has Boston turned into Baghdad?", a friend tweeted. The police had pursued the two bombing suspects into the neighboring suburb of Watertown where there had been a terrific firefight. One suspect died (died, as it turned out, because his brother ran him over) and his brother escaped. By the time we woke up in Cambridge, Watertown was already in lockdown and police were getting ready for house to house searches within a 20 block perimeter.

Then we heard that Boston and a few of its suburbs - roughly an area comprising a million inhabitants - were in lockdown and all residents had been asked to stay at home. I thought then and I still think that this was an overreaction. Watertown, where the suspect was thought to be hiding? Sure. But Boston, Cambridge, Belmont, Newton and four others? A little over the top in my opinion. I understand that many people stayed home out of deference to authorities' wishes to be able to do their job unfettered. It's also ok to ask residents to be vigilant and to venture out at their own risk, but we do this anyway. Every time we are out we run the risk of being in a traffic accident. I suspect that this risk in a random suburb which is not Watertown is probably higher than  a 19 year-old fanatic suffering blood loss coming out of the blue with guns blazing and shooting at you. Now I understand that the police did not force people to stay indoors but they were also quite emphatic about this; I watched a woman who stepped out in the middle of the day to walk her dog being emphatically told to stay inside by two officers out on patrol.

The huge police presence in Watertown also seemed like an overreaction to me. By one account there were 9000 local, state, and federal authorities looking for this kid. Armored vehicles patrolled the streets, and I am not sure what additional purpose they would have served. Sure, the authorities were erring on the side of safety and they were clearly anxious to apprehend the suspect as soon as possible, but I think it's constitutionally healthy to be skeptical when your whole neighborhood resembles a war zone and armed officers wielding every kind of weapon perform intrusive house searches.

For me the ultimate irony may be that this guy was located - not by one of the 9000 officers and military personnel - but by an ordinary citizen. In a boat in an area that was not part of the 20 block perimeter. After the lockdown order had been rescinded.

What happened there? I know that hindsight is always twenty-twenty but here's something that bothers me: From what I read it seems that the spectacular shootout occurred at the intersection of Laurel St and Dexter Ave in Watertown. The suspect was found hiding in the boat at 67 Franklin St. If you look at these locations on Google Maps they are less than a mile apart. For all the meticulous house-to-house searches and lockdowns, why did the perimeter not include a location that was less than a mile from where the shootout took place? And most importantly, how could the police miss the boat, a large, roomy object that's ideal for a human being to hide? Can you say that your operation was really successful when an ordinary citizen locates a suspect only after you are done with house-to-house searches? So on one hand there seemed to be an overreaction and on the other, the meticulous operation seems to have been unsuccessful in its primary purpose.

I understand that there were a lot of police officers and other personnel who immersed themselves into this investigation. Many of them had not slept in 24 hours and they were clearly committed to finding this guy as soon as possible. These people clearly did an admirable job and we should applaud their dedication. But in my opinion there seem to be a few important clues that were missed, and discussing these clues is not only an important part of a healthy democracy where public officials are answerable to the public but also a part of any system of self-improvement and feedback where you learn from your mistakes. Most importantly though, when a 19 year-old nutjob brings a major American city to a standstill, makes it resemble a state with martial law and makes people stay put in their houses and away from their jobs in anxiety, if not fear, the terrorists have already won (as the cliche goes, in this case because it's true). As Ben Franklin memorably put it, if you sacrifice freedom for security you risk losing both. And the key here is to realize that this sacrifice may not even be forced upon you by the state; it can be entirely self-imposed.

At 5 PM I grew really restless and decided to go outside to get some milk (I need my morning coffee fix, terrorist scares be damned). Everything except for one convenience store was closed. Parking on Massachusetts Ave never looked better. The next day we went to the Esplanade along the Charles River. The cherry blossoms were in full bloom. Something about fear being the only thing we should truly fear came to my mind.

Moore's Law for batteries: No dice

The REVAi/G-Wiz i electric car charging at an on-street station in London (Image: Wikipedia Commons)

Ever since Gordon Moore came up with the ubiquitous law bearing his name, it has been applied to paradigms far beyond those which it was intended for. This is perhaps not surprising; the history of science and technology - and of religion - has consistently demonstrated that the followers of a prophet usually extend his principles into domains which the prophet never really approved of.

Transistor technology does neatly seem to follow the Moore's Law curve and a few other cutting-edge technologies like genome sequencing also seem to do this. Yet Moore's proselytizers have extended his law to pretty much everything. The law especially seems to break down when applied to biomedical research; for instance a review from last year pointed out how the pace of drug development almost seems to have been following a reverse law, titled "Eroom's Law" of declining productivity. Kurzweilian prognostications notwithstanding, research in neuroscience might follow the same trajectory, with a burst of rapid mapping of neuronal connectivity followed by a long, fallow period in which we struggle to duplicate these processes by artificial means.

The basic reasons why an emerging technology may not follow Moore's Law is either because we tend to underestimate the complexity of the system to which the technology is applied, or we underestimate the basic principles of physics and chemistry which would inherently constrain a Moore-type breakthrough in that field. In case of medical research both these constraints seem to rear their ugly, emergent heads, and this is the main problem I have with futurists like Ray Kurzweil who seem to imagine an entire universe governed by Moore's Law-type exponential progress in every field. Not all levels of complexity are created equal, and we just don't have enough evidence to know how general Moore's Law (which I think should simply be re-named "Moore's Observation") is in the world of practical problem-solving.

The argument about basic science limitations may especially apply to much-touted battery research whose proponents often seem to declare the next breakthrough in battery technology as being just around the corner. But a perspective from Fred Schlachter from the American Physical Society in the Proceedings of the National Academy of Sciences puts a brake on these optimistic predictions. His point is simple: any kind of Moore's Law for batteries may be limited by the fundamental chemistry inherent in a battery's workings. This is unlike transistors, where finer lithography techniques have essentially enabled a repetitive application of miniaturization over the years.

There is no Moore’s Law for batteries. The reason there is a Moore’s Law for computer processors is that electrons are small and they do not take up space on a chip. Chip performance is limited by the lithography technology used to fabricate the chips; as lithography improves ever smaller features can be made on processors. Batteries are not like this. Ions, which transfer charge in batteries are large, and they take up space, as do anodes, cathodes, and electrolytes. A D-cell battery stores more energy than an AA-cell. Potentials in a battery are dictated by the relevant chemical reactions, thus limiting eventual battery performance. Significant improvement in battery capacity can only be made by changing to a different chemistry.

And even this different chemistry is going to be governed by fundamental parameters like the sizes of ions and the rates of chemical reactions and current flow. Schlachter goes on to note the problems that lithium batteries have recently encountered, including fires. There is thus no guarantee that there will be a breakthrough in battery technology that's equivalent to that in computer technology over the last thirty years. And the article is right that while we are waiting for such breakthroughs, it's a really good idea to push forward with improving energy efficiency in cars, making their lighter, smaller and and more powerful. Energy efficiency would not ultimately solve pollution problems since the cars would still be fueled by gasoline, but it would certainly take us a long way while we are waiting for the next battery breakthrough engineered by Moore's Law. A law which may not really hold when it comes to next generation electric technology.

First published on the Scientific American Blog Network.

Friday levity: 'Nature' discusses ghosts.

One of the pleasures of thumbing through old issues of science journals is the opportunity to accidentally discover articles or letters that make you do double takes, often followed by face palms. 

As I was about to read a letter in Nature bemoaning the closure of the Hoffmann-La Roche Institute of Molecular Biology in Nutley, NJ (Nature, 1995, 373, 184; deja vu, anyone?) I came across a letter on the same page that pristinely tosses out the following for readers' benefit (click for clarity...or the lack thereof).

I love the fact that the letter writer dismisses one hypothesis about ghosts only to come up with another. And this isn't 1885, it's 1995. Oh how I miss the old Nature.

Chemical compounds from mouthwash may target cancer cells

Apoptosis or programmed cell death is one of the great truths of cellular life, an essential process that’s not only required to make way for new cells but to prevent old cells from going haywire. When cells circumvent this great truth they start dividing uncontrollably and contribute to cancer. Our knowledge of cancer over the last three decades has confirmed the central role that a breakdown in the usual mechanisms of apoptosis plays in pushing a cell across the tipping point into a cancerous state. Of the many strategies to fight cancer, one consists of trying to find drugs that force cells to regain their normal balance of apoptosis. Now this effort may have found an unlikely ally.

Chlorhexidine is an antibacterial and plaque-fighting compound that is a common component of mouthwash, usually present as a 0.1% or 0.2% solution. In a paperpublished in the journal Angewandte Chemie, scientists in Germany report an unexpected effect of chlorhexidine and its related cousin alexidine: they inhibit cancer cells in the mouth by blocking an important protein-protein interaction. This research opens up new directions in investigating this class of compounds as anticancer agents and also sheds light on the value of finding novel potential uses for everyday chemical compounds. One of the great advantages in this endeavor is that the “repurposed” compounds have already run the gauntlet of safety tests required by the FDA, potentially shortening the period of approval for their new uses.

Protein-protein interactions (PPIs) are often considered the next frontier in drug discovery. They are involved in almost every important molecular-level event in health and disease. Traditional drugs work by blocking the action of single proteins (typically fitting into them like a key fits into a lock) but since there are many more protein-protein interactions than single proteins, there is enormous potential in developing drugs that disrupt these interactions, many of which are upregulated in diseases like cancer. Unfortunately targeting PPIs is difficult because of a variety of reasons; they have large, spread-out interfaces which makes it difficult for small organic molecules to span their surface area, and typically the ones which do are too big to satisfy the many qualities of an ideal drug, such as an ability to get inside cells in the first place.

One of the most well studied PPIs is the interaction between a family of pro-apoptotic and anti-apoptotic proteins called the Bcl-2 family. These proteins are present in all our cells. As their name indicates, one group of proteins speeds up apoptosis while the other group inhibits it. In a normal cell there is a usually a precise balance between these two activities engineered by the two sets of proteins binding to each other and regulating each other’s function. It’s a delicate dance which ensures that the cells are active only when needed and any cells gone haywire are eliminated. In cancer this precise balance is disrupted and the anti-apoptotic proteins are over-expressed and become dominant. One anti-apoptotic protein named Bcl-Xl in particular keeps its usually equipotent pro-apoptotic protein partner named Bak bound up and prevents the cell from committing suicide; this molecular-level feud leads to uncontrolled cell division. Over the years researchers have tried to find many druglike molecules and peptides which could block Bcl-Xl and free up the Bak protein. But none of the attempts have resulted in a clinically marketed drug.

What the researchers in Germany did was to screen about 4000 everyday chemical compounds to look for ones that might block the Bcl-Xl protein. They found two which, surprisingly, had very different uses. Chlorhexidine and alexidine are common components of mouthwash. Both compounds were found to inhibit the Bcl-Xl – Bak interaction at a concentration that’s much lower than that found in mouthwash. Surface-exposed oral cells in the mouth are thus bathed in a rather potent concentration of small molecules that prevent at least one important mechanism involved in cancer from manifesting itself. The researchers also did further experiments, including computer modeling, that localized the site of binding of the two compounds on the Bcl-Xl protein. This site was the same as that occupied by the Bak protein, further supporting the blocking interaction of the mouthwash components with the anti-apoptotic protein.

Finally the researchers tested these two compounds against cancerous cells from the tongue and the pharynx. Both compounds were found to significantly reduce the degree of apoptosis suppression in these cells, connecting the molecular level interaction of the molecules to actual anticancer effects.

This study is interesting for several reasons. It directly leads to a new class of compounds that may have promising anticancer activities; very likely the compounds’ structures would have to be modified by chemists to improve their properties, but this is what chemists have always done best. The therapeutic concentration that’s required for inhibiting the proteins is already exceeded in your garden variety mouthwash; this may also indicate a healthy margin of safety. A more intriguing question to ask is whether the use of mouthwash correlates with lower incidence of oral cancer. The literature on the relationship between mouthwash and oral cancer has been confusing and there don’t seem to be large-scale studies investigating a possible connection. By suggesting a possible mechanism of cancer prevention, this study provides a strong motivation to gather epidemiological data about possible anticancer effects of mouthwash and its components. It’s too early to start dousing your mouth with mouthwash though since these compounds only target one kind of interaction and we don’t have enough data on higher concentrations and long-term effects. But it’s definitely a promising start that points the way to interesting experiments, and that’s what science is best at doing.

Most tantalizingly though, the study asks what other kinds of therapeutic effects may be hidden in everyday chemical products, in our bathroom and kitchen closets. Nature is much more interesting than we think and molecules often lead double lives. Contemplate this the next time you brush your teeth or wash your dishes.

First published on the Scientific American Blog Network.

Solomon Snyder on academic publishing: ask for adequate, not exhaustive, documentation

Image: Corpus Callosum

Renowned neuropharmacologist Solomon Snyder has a thought-provoking take on what seems to be one of the two evils that has plagued modern academia: publication (the other one is the job market). I have previously blogged about the increasing conservatism of academic publishing myself, and in this case “conservatism” also translates to “excessive rigor”.

Snyder starts by lamenting the startling fact that the average duration for a modern American biomedical scientist to start his or her academic career is about the same as that for a neuro or cardiovascular surgeon, people whose specialty is usually considered to be in the top tier of their profession; the difference of course is that a cardiovascular surgeon starts making $500K right off the bat while a new assistant professor starts making $80K and almost never goes beyond $200k or so. The long trudge begins with graduate education, the average duration of which has stretched out over the last three decades (these days, a 5 year Ph.D. is considered relatively quick). Every part of the academic process, from getting a postdoctoral position to your first job to your first grant, has turned into a war of attrition. The “winners” who emerge at the end of it are often demoralized academics in their early 40s whose best years may be behind them. And the situation seems to only be getting worse.

But the article’s really about publishing papers. Snyder hits the nail when he says that academic publishing has become so rigorous in asking for exhaustive experimentation and documentation that it dissuades many authors from publishing their best ideas, ideas which are interesting and valid but which may not have been completely fleshed out. He points to reviewers’ insistence that authors perform a comprehensive set of experiments – often ranging over several months – that would qualify their manuscript for publication. Anyone who has tried to publish biomedical papers must be well aware of how tedious and demoralizing the experience can be. This long-drawn process significantly impacts the progress of science:

“Why does it take so much longer to move from test tube to the printed page? One element is a journal review process that is substantially lengthier, especially in terms of experiments required to address the concerns of referees. To anticipate such referee responses, scientists preemptively carry forward experimentation more exhaustively than is necessary to document their assertions. Yet, we can clone genes in a couple of days. Shouldn’t we be able to complete experiments to satisfy reviewers in a few weeks rather than the 7–12 months typically consumed in revision, not to mention the many years devoted to developing the original manuscript? If one spends 5 years accumulating the data for a manuscript and another year revising it to satisfy referees, benefits to the public are delayed for years.”

In contrast Snyder points to his postdoctoral advisor, another legendary scientist named Julius Axelrod at the NIH who churned out discovery after discovery in short order and won a Nobel Prize (the Axelrod dynasty is nicely charted out in Robert Kanigel’s book “Apprentice to Genius”). The point that Snyder is making is that in those days the reviewing process was much quicker but the quality of science doesn’t seem to have suffered in spite of this speedier turnaround. What has gone wrong since then?

Snyder partially places the blame at the feet of Cell founder Benjamin Lewin who wanted Cell to showcase papers that were essentially complete stories; from hypotheses to final products. But Lewin also made the process highly streamlined. Reviewers were warned to stay away from insults, stick to succinct criticism and suggest adequate but not unrealistic experiments and further studies. The objective was to get the best science out in a form that was interesting enough to spark further inquiry but which was not necessarily the last word.

Lewin understood the piecemeal nature of science where researchers build on each other’s discoveries. This understanding of the scientific process has since been subverted by academic reviewers, partially to cull a flood of proposals and ideas and partially to satisfy their own whims. Sometimes old boys’ networks can conspire to put sound science in a straitjacket. Expecting every research project to tell complete, final stories not only imposes unrealistic and demotivating standards on scientists but also ignores the always incomplete and provisional nature of science. Snyder asks that expectations for accepting papers be changed and points to recent developments like the journal eLIFE which incorporates some of his thinking. Blogger SciCurious suggests her own system of peer-review where a paper is simultaneously sent to a group of journals with different standards; after hearing back from reviewers, the authors can decide whether to push ahead with further experiments to satisfy the top-tier journals or whether to publish the paper in a lower-tier journal right away. But Snyder’s perspective points out that all journals – whether top tier or otherwise – should have a reviewing system that allows for rapid dissemination of results.

Reviews and authors need to seriously contemplate Snyder’s recommendations. Academic research has already turned into a long slog with its uncertain job market and draconian grant approval and does not to face need additional difficulties in the form of glacial and unrealistic reviewing standards. Let’s remember that the purpose of science is to generate ideas, not products. And it shouldn’t take very long for ideas to see the light of day.

First published on Scientific American Blogs.

First they came for the bloggers and I didn't speak up because...

Here's a breath of fresh air. I keep on thinking about Planck's quote about scientific revolutions not occurring until old generations die and new ones take their place and here's something of that sort happening, even if in a minor way. 

Prof. Phil Baran has started his own blog. It's easy to see this as a response to the commendable IBX oxidation experiment carried out by Blog Syn. That, by the way, was a great illustration of how science should work; research is published, it is then scrutinized, a few discrepancies are found, the original author responds and confirms the original results, and the new authors discover something new that had not been realized before.

But it's clear that the Blog Syn incident was only a seed for a realization that undoubtedly must have been crystallized in Phil's mind for a while. We have all seen how the old guard has often dismissed and scorned bloggers and their pesky, amateur blogs. Now here's someone from the new guard who clearly recognizes which way the winds are blowing:

Over the years I have vaguely followed some of them, mostly through my students or through being occasionally contacted by someone that runs a blog. Practically all of my colleagues roll their eyes the minute the word "blog" is uttered for a variety of largely justified reasons.  

But times are clearly changing...Last year I was at a dinner symposium where EJ Corey gave a brilliant impromptu talk before a toast. It was a captivating speech all about how things have rapidly changed over the span of his 80+ years. The take home message was that change is natural and you can either embrace it and adapt or be left behind. I'm no fortune teller but it is clear to me that blogging is here to stay and is gathering momentum.

There is something ironic about the fact that the words of the great E. J. Corey - an exemplar of the old guard who almost certainly is not going to start blogging anytime soon - should serve as an invitation to blog for the man who is widely regarded as one of the most creative synthetic organic chemists of his generation. 

Thank you Prof. Baran. Now, the next time they come for us, we know you will speak up because you are are a blogger.

Other reactions: Chembark

ENCODE, Apple Maps and function: Why definitions matter

ENCODE (Image: Discover Blogs)

Remember that news-making ENCODE study with its claims that “80% of the genome is functional”? Remember how those claims were the starting point for a public relations disaster which pronounced (for the umpteenth time) the "death of junk DNA"? Even mainstream journalists bought into this misleading claim. I wrote a post on ENCODE where I expressed surprise at why anyone would be surprised by junk DNA to begin with.

Now Dan Graur and his co-workers from the University of Houston have published a meticulous critique of the entire set of interpretations from ENCODE. Actually let me rephrase that. Dan Graur and his co-workers have published a devastating takedown of ENCODE in which they pick apart ENCODE’s claims with the tenacity and aplomb of a vulture picking apart a wildebeest carcass. Anyone who is interested in ENCODE should read this paper, and it’s thankfully free.

First let me comment a bit on the style of the paper which is slightly different from that in your garden variety sleep-inducing technical article. The title – On the Immortality of Television Sets: Function in the Human Genome According to the Evolution-Free Gospel of ENCODE – makes it clear that the authors are pulling no punches, and this impression carries over into the rest of the article. The language in the paper is peppered with targeted sarcasm, digs at Apple (the ENCODE results are compared to AppleMaps), a paean to Robert Ludlum and an appeal to an ENCODE scientist to play the protagonist in a movie named "The Encode Incongruity". And we are just getting warmed up here. The authors spare little expense in telling us what they think about ENCODE, often using colorful language. Let me just say that if half of all papers were this entertainingly written, the scientific literature would be so much more accessible to the general public.

On to the content now. The gist of the article is to pick apart the extremely liberal, misleading and scarcely useful definition of “functional” that the ENCODE group has used. The paper starts by pointing out the distinction between function that’s selected for and function that’s merely causal. The former definition is evolutionary (in terms of conferring a useful survival advantage) while the latter is not. As a useful illustration, the function of the human heart that is selected for is to pump blood while the function that’s causal is an additional weight of 300 grams and a capacity for producing thumping sounds.

The problem with the ENCODE data is that it features causal functions, not selected ones. Thus for instance, ENCODE assigns function to any DNA sequence that displays a reproducible signature like binding to a transcription factor protein. As this paper points out, this definition is just too liberal and often flawed. For instance a DNA sequence may bind to a transcription factor without inducing transcription. In fact the paper asks why the study singled out transcription as a function: “But, what about DNA polymerase and DNA replication? Why make a big fuss about 74.7% of the genome that is transcribed, and yet ignore the fact that 100% of the genome takes part in a strikingly “reproducible biochemical signature” – it replicates!”

Indeed, one of the major problems with the ENCODE study seems to be its emphasis on transcription as a central determinant of “function”. This is problematic, since as the authors note, there's lots of sequences that are transcribed which are known to have no function. But before we move on to this, it’s worth highlighting what the authors call “The Encode Incongruity” in homage to Robert Ludlum. The Encode Incongruity points to an important assumption in the study; the implication that a biological function can be maintained without selection and that the sequences with “causal function” identified by ENCODE will not accumulate deleterious mutations. This assumption is unjustified.

The paper then revisits the five central criteria used by ENCODE to define “function” and carefully takes them apart:

1. “Function” as transcription.

This is perhaps the biggest bee in the bonnet. First of all, it seems that ENCODE used pluripotent stem cells and cancer cells for its core studies. The problem with these cells is that they display a much higher level of transcription than other cells, so any deduction of function from transcription in these cells would be exaggerated to begin with. But more importantly as the article explains, we already know that there are three classes of sequences that are transcribed without function; introns, pseudogenes and mobile elements (“jumping genes”). Pseudogenes are an especially interesting example since they are known to be inactive copies of protein-coding genes that have been rendered dead by mutation. Over the past few years as experiments and computational algorithms have annotated more and more genes, the number of pseudogenes has gone up even as the number of protein-coding genes has gone down. We also know that pseudogenes can be transcribed and even translated in some cells, especially of the kind used in ENCODE, just as we know that they are non-functional by definition. Similar arguments apply to introns and mobile elements, and the article cites papers which demonstrate that knocking these genes out doesn't impair function. So why would any study label these three classes of sequences as functional just because they are transcribed? This seems to be a central flaw in ENCODE.

A related point made by the authors is statistical in which they say that the ENCODE project has sacrificed selectivity for sensitivity. There are some simple numerical arguments that point to the large number of false positives inherent in sacrificing selectivity for sensitivity. In fact this is a criticism that goes to the heart of the whole purpose of the ENCODE study:

“At this point, we must ask ourselves, what is the aim of ENCODE: Is it to identify every possible functional element at the expense of increasing the number of elements that are falsely identified as functional? Or is it to create a list of functional elements that is as free of false positives as possible. If the former, then sensitivity should be favored over selectivity; if the latter then selectivity should be favored over sensitivity. ENCODE chose to bias its results by excessively favoring sensitivity over specificity. In fact, they could have saved millions of dollars and many thousands of research hours by ignoring selectivity altogether, and proclaiming a priori that 100% of the genome is functional. Not one functional element would have been missed by using this procedure.”

2. “Function” as histone modification

Histones are proteins that pack DNA into chromatin. The histones then undergo certain chemical modifications called post-translational modifications that cause the DNA to unpack and be expressed. ENCODE used the presence of 12 histone modifications as evidence of “function”. This paper cites a study that found a very small proportion of possible histone modifications associated with function. Personally I think this is an evolving area of research but I too question the assumption of having a function associated with most histone modifications.

3. “Function” as proximity to regions of open chromatin

In contrast to histone-packaged DNA, open chromatin regions are not bound by histones. ENCODE found that 80% of transcription sites were within open chromatin regions. But then they seem to have committed the classic logical fallacy of inferring the opposite, that most open chromatin regions are functional transcription sites (there’s that association between transcription and function again). As the authors note, only 30% or so of open chromatin sites are even in the neighborhood of transcription sites, so associating most open chromatin sites with transcription seems to be a big leap to say the least.

4. “Function” as transcription-factor binding.

This to me is another huge assumption inherent in the ENCODE study, especially as a chemist. As I mentioned in my earlier post, there are regions of DNA that might bind transcription factors (TFs) just by chance through a few weak chemical interactions. The binding might be extremely weak and may be a quick association-dissociation event. To me it seemed that in associating any kind of transcription-factor binding with function, the ENCODE team had inferred biology from chemistry. The current analysis gives voice to my suspicions. As the authors say, transcription sites are usually very short which means that TF-binding “look-alikes” may arise in a large genome purely by chance. Any binding to these sites may be confused with real TF-binding sites. The authors also cite a study in which only 86% of TF-binding sites in a small sample of 14 sites showed experimental binding to a TF. Extrapolating to the entire genome, it could mean that a fraction of the conjectured TF-binding sites may actually bind TFs.

5. “Function” as DNA methylation.

This is another instance in which it seems to me that biology is being inferred from chemistry. DNA methylation is one of the dominant mechanisms of epigenetics. But by itself DNA methylation is only a chemical reaction. The ENCODE team built on a finding that negatively correlated gene expression with methylation in CpG (cytosine-guanine) sites.  Based on this they concluded that 96% of all CpGs in the genome are methylated, and therefore functional. But again, in the absence of explicit experimental verification, CpG methylation cannot be equated with gene expression. At the very least this indicates follow-up work which will need to confirm the relationship. Until then the hypothesis that CpG methylation implies function will have to remain a hypothesis.

So what do we make of all this? It’s clear that many of the conclusions from ENCODE have been extrapolations devoid of hard evidence. 

But the real fly in the ointment is the idea of “junk DNA” which seems to have evoked rather extreme opinions that have ranged from proclaiming junk DNA as extinct to proclaiming it as God. Both these opinions perform a great disservice to the true nature of the genome. The former reaction virtually rolls the red carpet for “designer” creationists who can now enthusiastically remind us of how each and every base pair in the genome has been lovingly designed. At the same time, asserting that junk DNA must be God is tantamount to declaring that every piece of currently designated junk DNA must forever be non-functional. While the former transgression is much worse, it’s important to amend the latter belief. To do this the authors remind us of a distinction made by Sydney Brenner between “junk DNA” and “garbage DNA”. There’s the rubbish we keep and the rubbish we discard, but some rubbish may potentially turn useful in the future. At the same time, rubbish that may be useful in the future is not rubbish that’s useful in the present. Just because some “junk DNA” may turn out to have a function in the future does not mean most junk DNA will be functional. In fact as I mentioned in my post, the presence of large swathes of non-functional DNA in our genomes is perfectly consistent with standard evolutionary arguments.

The paper ends with an interesting discussion about “small” and “big” science that may explain some of the errors in the ENCODE study. The authors point out that big science has generally been in the business of generating and delivering data in an easy-to-access format. Small science has been much more competent in then interpreting the data. This does not mean that scientists working on big science are incapable of data interpretation; what it means is that the very nature of big data (and the time and resource allocation inherent in it) may make it very difficult for these scientists to launch the kinds of targeted projects that would do the job of careful data interpretation. Perhaps, the paper suggests, ENCODE’s mistake was in trying to act as both the deliverer and the interpreter of data. In the authors’ considered opinion, ENCODE “tried to perform a kind of textual hermeneutics on the 3.5 billion base-pair genome, disregarded the rules of scientific interpretation and adopted a position of theological hermeneutics, whereby every letter in a text is assumed a priori to have a meaning”. In other words, ENCODE seems to have succumbed to an unfortunate case of ubiquitous pattern seeking from which humans often suffer.

In any case, there are valuable lessons in this whole episode. The mountains of misleading publicity it generated, even in journals like Science and Nature, were a textbook study in media hype. As the authors say:

“The ENCODE results were predicted by one of its lead authors to necessitate the rewriting of textbooks (Pennisi 2012). We agree, many textbooks dealing with marketing, mass-media hype, and public relations may well have to be rewritten.”

From a scientific viewpoint, the biggest lesson here may be to always keep fundamental evolutionary principles in mind when interpreting large amounts of noisy biological data under controlled laboratory conditions. It’s worth remembering the last line of the paper:

“Evolutionary conservation may be frustratingly silent on the nature of the functions it highlights, but progress in understanding the functional significance of DNA sequences can only be achieved by not ignoring evolutionary principles…Those involved in Big Science will do well to remember the depressingly true popular maxim: “If it is too good to be true, it is too good to be true.”

The authors compare ENCODE to AppleMaps, the direction-finding app in the iPhone that notoriously bombed when it came out. Yet AppleMaps also provides a useful metaphor. Software can evolve into a useful state. Hopefully, so will our understanding of the genome.

First published on the Scientific American Blog Network.

Uncle Syd's idea for funding new assistant professors

I was leafing (virtually of course) through old issues of "Current Biology" when I came across a thought-provoking, slightly tongue-in-cheek essay by Sydney Brenner in an issue from 1994. Brenner used to write a regular column for the magazine and his thoughts ranged from improving lab conditions to junk DNA; as is characteristic of Brenner's incisive mind, almost all the columns give you something new to think about. 

It was Brenner's ("Uncle Syd") fictitious letter to an assistant professor ("Dear Willie") just starting out in his new career that caught my eye. Perhaps the column can offer some ideas to assistant professors floundering in this gloomy age of funding crunches and declining job prospects. After acknowledging the fundamental and paradoxical difficulty that new professors who need funding the most don't have any experience in getting it, Brenner comes up with a framework - the BISCUIT:

"I have for long entertained an elegant solution to this difficulty, and that is to found a bank, BISCUIT (Bank of International Scientific Capital and Unpublished Information and Techniques), that will lend scientific capital to first-time grant applicants and others in need. It will not only lend ideas for research but also loan experiments that have been carried out but have not been published. We have to be careful with the latter, because although such holdings are of high value they could undergo instant depreciation if someone else does the experiment and publishes the result. Where, you ask, does the bank get its capital? No problem. I know a number ofscientists who have a surplus of scientific ideas and lots of experiments that they find too boring to write up and these 'wealthy' individuals would be the first investors. The bank would also continue to receive deposits. Once we got going, everything would be fine, because the borrowers would not only have to pay back capital but we would charge interest so that our holdings grew. And, of course, if any depositor were to suffer a catastrophic career collapse, he could withdraw all of his capital and start again. The beauty of it is that he would get new, up-to-date ideas and experiments and, in this way, his original deposit, although used a long time ago, will have retained its value and will not have been corroded by time. I am amazed that in these times ofhigh-powered service industries nobody has thought of doing this before, but perhaps that's because it is only scientists who will profit from the BISCUIT bank."

Reference: Current Biology, 1994, 4, 10, 956

On toxic couches and carcinogens: Chemophobia, deconstructed.

Last week I attended a great session on chemophobia at ScienceOnline 2013 headed by Carmen Drahl and Dr. Rubidium. The session emphasized how "trigger words" - alarmist phrases judiciously placed in the middle of otherwise well-intentioned paragraphs - can make people believe that something is more serious than it is. The session also reinforced the all-important point that context makes all the difference when it comes to chemistry.

Sadly I could not read a recent post about flame retardants in couches on the Scientific American Guest Blog without remembering some of these caveats. The post unfortunately seems to me to present a first-rate example of how well-intentioned opinion and advice can nonetheless be couched in alarmism and assertions drawn out of context. It evidences lapses that are common in chemophobic reporting. Let me state upfront that my argument is as much about the tone and message of the post as it is with the pros and cons of the scientific evidence (although there's some highly questionable scientific conclusions in there). Some of my analysis might look like nitpicking, but the devil is often in the details.

The article is written by Sarah Janssen, an M.D. Ph.D. who is worried about supposedly "toxic chemicals" in her couch. In this case the chemical turns out to be something called Chlorinated Tris. Dr. Janssen is apparently so worried that she has already decided that her family should sit on the floor/carpet or eat at the table than be exposed to the couch. At the end of the post she says that she will look forward to the time when she can buy a "toxic-free couch".

The trigger words start coming at you pretty much right away:

"A nationwide study of 102 couches revealed that my couch, among others tested, contains OVER A POUND of chlorinated Tris, a cancer-causing chemical removed from children’s pajamas in the 1970s and now listed on California’s Proposition 65 list of carcinogens"

Observe how ONE POUND is capitalized, as if the capitalization makes any additional arguments in favor of the compound's toxicity superfluous. But we all know that the dose depends on the context; there's more than one pound of lots of chemical substances in almost every piece of furniture that I use, but the weight by itself hardly makes the material harmful. In fact since the weight of a typical couch is at least 20 pounds, I wouldn't expect to find any less than one pound of a flame-retardant substance in it. The point is that simple manipulations like capitalization enhance the public's perception of impact, and doing this without a good reason sends the wrong message.

Now let's look at the chemical itself, Chlorinated Tris, or TRIS(1,3-DICHLORO-2-PROPYL) PHOSPHATE (TDCPP) in chemical parlance (there, did the capitalization make it sound more sinister?). Googling this chemical turns up a bunch of newspaper articles without primary references. How about a more formal source, in this case a June 2011 report by the California EPA? Scientifically inclined readers will find lots of interesting data in there and it's clear that chlorinated tris has a variety of observable and potentially concerning effects on cells. But for me the most important part of the report talked about a study on the effects of TDCPP on cancer risk in a group of 289 workers at a TDCPP plant between 1956 to 1980. The operative line in that paragraph is the following:

"The authors concluded that although the SMR (standard mortality rate) from lung cancer was higher than expected, overall there was no evidence linking the lung cancers to TDCPP exposure because all three cases with lung cancer were heavy to moderate cigarette smokers. Small sample size and the inability to account for confounding factors make it difficult to draw conclusions from this study."

In addition the paragraph states that p-values (a measure of statistical significance) could not be calculated because of small sample size. Now this study was done with people who have literally lived and breathed in a TDCPP-rich environment for almost thirty years. If anyone should suffer the ill-effects of TDCPP it should probably be this group. And still the conclusions were dubious at best, so one wonders if merely sitting on a couch would do anything at all.

As is usually the case, the report has much more information about the effects of TDCPP in mice and here you do see evidence of tumor formation. But the sample sizes are again small. More importantly, what's the dosage of TDCPP that causes statistically significant cancers to appear in mice? It's 80 mg per kilogram per day. This would translate to 5.6 grams per day for a 70-kg human being. And although I haven't read all the original studies with mice, I am assuming that this amount would have to be ingested, inhaled or injected. So no, unless you are out of supplies in a nuclear holocaust and are forced to survive by actually eating the foam from your couch, you would most likely not get cancer from simply sitting on a couch with TDCPP in it. And even this tenuous conclusion comes from studies with mice; as indicated above, the data is far from clear for humans. In fact I would guess that the probability of suffering an obesity-induced heart attack from sitting for long periods on a couch exceeds the probability of getting cancer from TDCPP.

Now that doesn't mean that I am claiming that TDCPP has no harmful effects in humans. But it's clear that at the very least we need to get much more rigorous data to establish a causal relationship with any kind of confidence. For now the evidence just doesn't seem to be there. Claiming that simply sitting on a TDCPP-filled couch could cause cancer, with any kind of probability, is really no more than a theory disconnected from data.

It gets worse. The post later talks about the smoke from fire retardant-containing furniture "putting firefighters' health at greater risk of cancer". When you click on that link it takes you to an article in the San Francisco Chronicle documenting the story of a firefighter named Stefani who is a cancer survivor. The firefighter had expressed concerns about his cancer being linked to smoke inhalation from household items like furniture. But here's what the article itself says at one point:

"The relationship between Stefani's job and transitional cell carcinoma is less clear...there's no hard evidence yet that chemicals contribute to this condition, said Dr. Kirsten Greene, a UCSF assistant urology professor who helped run the study."

When an expert who ran the study questions the link between cancer and flame retardants, it should give you pause for thought (on a related note, kudos to the SFC for reporting the skepticism).

Finally, there's no better way to drive home the pernicious influence of "chemicals" than to demonstrate their existence in the bodies of every species on the planet:

"But flame retardants aren’t just polluting our homes—they are polluting the world, literally. During manufacturing, use and disposal, these chemicals are released into the environment where they can be found in air, water, and wildlife. Birds, fish, mammals including whales and dolphins and animals living far from sources of exposure, such as polar bears in the Arctic, have been found to have flame retardants in their bodies."

But take a look at the polar bears paper. Notice that we have now switched from TDCPP to brominated flame retardants, a different category of compound. If there's anything chemists know, it's the fact that function follows structure; no chemist would assume that TDCPP and brominated ethers would have the same effects without explicit evidence. The subsequent paragraph describing a variety of other non-carcinogenic effects also talks about brominated compounds. The continuity in the article would have you believe that we are still talking about TDCPP and couches. More importantly though, chemical substances are not all toxic just because they show up in multiple species, and bioaccumulation does not automatically translate into carcinogenicity (as is clear from the polar bear study). Since the advent of human civilization there have been thousands of synthetic molecules that have been dispersed in the environment, and the vast majority of them co-exist in peace with other species. But the most important point here is that it's extremely hard to extrapolate these studies to the conclusion that sitting on your couch may expose you to a carcinogen; that kind of extrapolation pretty much ignores dose, context, statistical significance and species-specific differences and lumps all "flame retardants" into the same category without allowing for compound-specific effects.

Trigger words proliferate the rest of the post: "dangerous chemicals", "harmful chemical substances", "toxic-free couch"...the list goes on. I don't want it to sound like I am picking on this particular post or author; sadly this kind of context-free alarmism is all too common in our chemophobic culture. But articles like this keep on making one thing clear: the details matter. You really cannot write a report like this without looking into details like statistics, nature of test organisms, dosage, method of administration, controls, sample size and species-specific differences. Lack of attention to these details is often a common hallmark of articles propagating chemophobia. If you ignore these details you are not really reporting science, you are simply reporting a gut feeling. And gut feelings are not exactly good metrics for making policy decisions.