Genetic architecture of intelligence (video with improved sound)

This is video of a talk given to cognitive scientists about a year ago (slides). The original video had poor sound quality, but I think I've improved it using Google's video editing tool. I found that the earlier version was OK if you used headphones, but I think they are not necessary for this one.

I don't have any update on the BGI Cognitive Genomics study. We are currently held up by the transition from Illumina to Complete Genomics technology. (Complete Genomics was acquired by BGI last year and they are moving from the Illumina platform to an improved CG platform.) This is highly frustrating but I will just leave it at that.

On one of the last slides I mention Compressed Sensing (L1 penalized convex optimization) as a method for extracting genomic information from GWAS data. The relevant results are in this paper.

Hints of genomic dark matter: rare variants contribute to schizophrenia risk

The study below is sensitive to rare variants which are implicated in schizophrenia risk. These rare variants add to the heritability already associated with common variants, estimated to be at least 32%. In related work, mutations affecting schizophrenia risk were shown to depress IQ in individuals who did not present for schizophrenia.

These results suggest a model for individual variation in brain function and cognitive ability driven by the number of disruptive mutations ("nicks" affecting key functions such as those listed below). Some of these variants are relatively common in the population (e.g., have frequency of at least a few percent), others are very rare (e.g., 1 in 10,000 in the population). See also Deleterious variants affecting traits that have been under selection are rare and of small effect.

A polygenic burden of rare disruptive mutations in schizophrenia
doi:10.1038/nature12975

Schizophrenia is a common disease with a complex aetiology, probably involving multiple and heterogeneous genetic factors. Here, by analysing the exome sequences of 2,536 schizophrenia cases and 2,543 controls, we demonstrate a polygenic burden primarily arising from rare (less than 1 in 10,000), disruptive mutations distributed across many genes. Particularly enriched gene sets include the voltage-gated calcium ion channel and the signalling complex formed by the activity-regulated cytoskeleton-associated scaffold protein (ARC) of the postsynaptic density, sets previously implicated by genome-wide association and copy-number variation studies. Similar to reports in autism, targets of the fragile X mental retardation protein (FMRP, product of FMR1) are enriched for case mutations. No individual gene-based test achieves significance after correction for multiple testing and we do not detect any alleles of moderately low frequency (approximately 0.5 to 1 per cent) and moderately large effect. Taken together, these data suggest that population-based exome sequencing can discover risk alleles and complements established gene-mapping paradigms in neuropsychiatric disease.

Figure from related paper: De novo mutations in schizophrenia implicate synaptic networks (doi:10.1038/nature12929)

Caption: Overlap of genes bearing nonsynonymous (NS) de novo mutations in schizophrenia (refs 12–14), autism spectrum disorder (refs 6–9) and intellectual disability (refs 10, 11). Overlaps of six or fewer genes are listed by name. See Extended Data Table 5 for statistical significance of these overlaps, and see Table 2 and text for disease sets.

Scholastic achievement by SES over four decades

This paper (click link below) has data on SAT, HSGPA, level of HS math and science completed, etc. from 1972 to 2004, binned by SES.

Running in Place: Low-Income Students and the Dynamics of Higher Education Stratification

The increasing concentration of wealthy students at highly selective colleges is widely perceived, but few analyses examine the underlying dynamics of higher education stratification over time. To examine these dynamics, the authors build an analysis data set of four cohorts from 1972 to 2004. They find that low-income students have made substantial gains in their academic course achievements since the 1970s. Nonetheless, wealthier students have made even stronger gains in achievement over the same period, in both courses and test scores, ensuring a competitive advantage in the market for selective college admissions. Thus, even if low-income students were “perfectly matched” to institutions consistent with their academic achievements, the stratification order would remain largely unchanged. The authors consider organizational and policy interventions that may reverse these trends.  [ Italics mine ]

The Presidents

President Obama toured the Michigan Biotechnology Institute today before signing the Farm Bill on the MSU campus. I'll post a photo of us together when I get it from the White House photographer.

I had a brief chat with the soldier carrying the football, and discussed Crossfit with the secret service guys. Favorite film: To Live and Die in LA :-)

MSU President Simon: Steve's a physicist who also knows genomics. He's doing great things here.

President Obama: He looks like someone who can do great things!

Labor Intensive or Labor Expensive?

New report on higher education costs and staffing trends based on Delta Cost Project data, including IPEDS input. Also discussed at the Chronicle.

... This report looks at long-term employment changes on college and university campuses during the past two decades and examines fluctuations in faculty staffing patterns, growth in administrative positions, and the effects of the recent recession on long-standing employment trends. It goes beyond other studies (Zaback, 2011; Bennett, 2009) to explore the effects of these staffing changes on total compensation, institutional spending patterns, and ultimately tuitions.

The overarching trends show that between 2000 and 2012, the public and private nonprofit higher education workforce grew by 28 percent, more than 50 percent faster than the previous decade. But the proportion of staff to students at public institutions grew slower in the 2000s than in the 1990s because the recent expansion in new positions largely mirrored rising enrollments as the Millennial Generation entered college. By 2012, public research universities and community colleges employed 16 fewer staff per 1,000 full-time equivalent (FTE) students compared with 2000, while the number of staff per student at public master’s and bachelor’s colleges remained unchanged.

... Growth in administrative jobs was widespread across higher education — but creating new professional positions, rather than executive and managerial positions, is what drove the increase. Professional positions (for example, business analysts, human resources staff, and admissions staff) grew twice as fast as executive and managerial positions at public nonresearch institutions between 2000 and 2012, and outpaced enrollment growth.

Colleges and universities have invested in professional jobs that provide noninstructional student services, not just business support. Across all educational sectors, wage and salary expenditures for student services (per FTE staff) were the fastest growing salary expense in many types of institutions between 2002 and 2012.

Part-time faculty/graduate assistants typically account for at least half of the instructional staff in most higher education sectors. Institutions have continued to hire full-time faculty, but at a pace that either equaled or lagged behind student enrollments; these new hires also were likely to fill non-tenure-track positions.

Part-time faculty (and graduate assistants) provided additional capacity at well-funded research universities and private colleges, but replaced new, full-time positions at broadly accessible, public master’s and bachelor’s institutions.

As the ranks of managerial and professional administrative workers grew, the number of faculty and staff per administrator continued to decline. The average number of faculty and staff per administrator declined by roughly 40 percent in most types of four-year colleges and universities between 1990 and 2012, and now averages 2.5 or fewer faculty and staff per administrator.

PGD in the US and Israel

Two recent articles in the Times on Preimplantation Genetic Diagnosis, or PGD. The first article focuses on ethical issues and a couple in the US, while the second follows the fertility travails of a Times writer in Israel, where PGD and IVF are covered by national health insurance.

NYTimes: ... Genetic testing of embryos has been around for more than a decade, but use of the procedure has soared in recent years as methods have improved and more disease-causing genes have been discovered. The in vitro fertilization and testing are expensive — typically about $20,000 — but they make it possible for couples to ensure that their children will not inherit a faulty gene and to avoid the difficult choice of whether to abort a pregnancy if testing of a fetus detects a genetic problem.

But the procedure also raises unsettling ethical questions that trouble advocates for the disabled and have left some doctors struggling with what they should tell their patients. When are prospective parents justified in discarding embryos? Is it acceptable, for example, for diseases like GSS, that develop in adulthood? What if a gene only increases the risk of a disease? ...

There is no question that the method’s use is increasing rapidly, though no group collects comprehensive data, said Dr. Joe Leigh Simpson, vice president for research at the March of Dimes and past president of the American Society for Reproductive Medicine.

The center in Chicago that tested the Kalinskys’ embryos for the GSS gene, the Reproductive Genetics Institute, has tested embryos from more than 2,500 couples and sought to identify 425 gene mutations. The institute’s caseload has risen 40 percent in just the past two years, according to its director, Svetlana Rechitsky, an author of the new paper on the Kalinskys’ case.

Ms. Kalinsky’s gene causes a particularly dire disease. Her grandfather, great-aunt, uncle, father and cousins died of it. Sometime between her mid-30s and her mid-50s, Ms. Kalinsky, who is now 30, will begin to stumble like a drunk. Dementia will follow, and possibly blindness or deafness. Five years after the onset of symptoms, she will most likely be dead. ...

Meanwhile, in Israel:

NYTimes: ... We knew what treatment we wanted: full-force I.V.F. using Preimplantation Genetic Diagnosis or Screening (P.G.D./P.G.S.), which genetically screens embryos before implantation. The method can be used to search for specific diseases such as Tay Sachs or Gaucher, but it can also weed out aneuploidy (abnormal number of chromosomes in a cell), the reason for many first-trimester miscarriages.

Some worry about the use of P.G.D. for gender selection (“Will this become a form of eugenics?”). But I wonder if genetic screening will be the future of reproductive medicine, especially since new technologies keep emerging. Then again, reproductive medicine is still in its nascent stages, so who knows what the future will bring? Maybe all women will freeze their eggs at age 25. Maybe no one will ever have fertility trouble again.

Genetic screening may not increase my chances of having a baby, but it would greatly reduce my chances of miscarrying or bearing a baby with genetic defects. After three miscarriages, it seemed like the best option: I couldn’t put myself through another such trauma. Besides, if all our embryos were bad, we would definitively know it was time to move on.

Embryo screening costs at least $5,000 for one to eight embryos in the United States, but in Israel it would be covered by national health care. ...

Genetic engineering of monkeys using CRISPR

For earlier CRISPR posts, see here. Off-target mutations seem to be under control in this study. Technology Review.

Generation of Gene-Modified Cynomolgus Monkey via Cas9/RNA-Mediated Gene Targeting in One-Cell Embryos (DOI:10.1016/j.cell.2014.01.027)

Monkeys serve as important model species for studying human diseases and developing therapeutic strategies, yet the application of monkeys in biomedical researches has been significantly hindered by the difficulties in producing animals genetically modified at the desired target sites. Here, we first applied the CRISPR/Cas9 system, a versatile tool for editing the genes of different organisms, to target monkey genomes. By coinjection of Cas9 mRNA and sgRNAs into one-cell-stage embryos, we successfully achieve precise gene targeting in cynomolgus monkeys. We also show that this system enables simultaneous disruption of two target genes (Ppar-γ and Rag1) in one step, and no off-target mutagenesis was detected by comprehensive analysis. Thus, coinjection of one-cell-stage embryos with Cas9 mRNA and sgRNAs is an efficient and reliable approach for gene-modified cynomolgus monkey generation.

In other exciting news: Bidirectional developmental potential in reprogrammed cells with acquired pluripotency (DOI:10.1038/nature12969): stimulus-triggered acquisition of pluripotency (STAP).

China 1793

Amasa Delano on China in the late 1700s, from his Narrative of Voyages and Travels in the Northern and Southern Hemispheres: Comprising Three Voyages Round the World; Together with a Voyage of Survey and Discovery, in the Pacific Ocean and Oriental Islands (full text via Google Books; p. 531 and following). Delano was in Canton while Macartney's embassy took place in Beijing. 

See also Koxinga for a comparison of European and Chinese development in this era. 

In the ensuing century both France and England would adopt examination systems, following the Chinese model, for civil service and university selection.

Locality and Nonlinear Quantum Mechanics

New paper!

Locality and Nonlinear Quantum Mechanics
(http://arxiv.org/abs/1401.7018)

Chiu Man Ho, Stephen D.H. Hsu

Nonlinear modifications of quantum mechanics generically lead to nonlocal effects which violate relativistic causality. We study these effects using the functional Schrodinger equation for quantum fields and identify a type of nonlocality which causes nearly instantaneous entanglement of spacelike separated systems.

Some excerpts:

The linear structure of quantum mechanics has deep and important consequences, such as the behavior of superpositions. One is naturally led to ask whether this linearity is fundamental, or merely an approximation: Are there nonlinear terms in the Schrodinger equation?

Nonlinear quantum mechanics has been explored in [1–6]. It has been observed that the fictitious violation of locality in the Einstein-Podolsky-Rosen (EPR) experiment in conventional linear quantum mechanics might become a true violation due to nonlinear effects [7, 8] (in [8] signaling between Everett branches is also discussed). This might allow superluminal communication and violate relativistic causality. These issues have subsequently been widely discussed [9].

Properties such as locality or causality are difficult to define in non-relativistic quantum mechanics (which often includes, for example, “instantaneous” potentials such as the Coulomb potential). Therefore, it is natural to adopt the framework of quantum field theory: Lorentz invariant quantum field theories are known to describe local physics with relativistic causality (influences propagate only within the light cone), making violations of these properties easier to identify. ...

... Our results suggest that nonlinearity in quantum mechanics is associated with violation of relativistic causality. We gave a formulation in terms of factorized (unentangled) wavefunctions describing spacelike separated systems. Nonlinearity seems to create almost instantaneous entanglement of the two systems, no matter how far apart. Perhaps our results are related to what Weinberg [11] meant when he wrote “... I could not find any way to extend the nonlinear version of quantum mechanics to theories based on Einstein’s special theory of relativity ... At least for the present I have given up on the problem: I simply do not know how to change quantum mechanics by a small amount without wrecking it altogether.”

UFC training in Ko Samui

from THE REEM on Vimeo.

Great documentary of Overeem fighting in Boston and training in Ko Samui. The locations look very familiar to me even though I was last in Thailand almost 20 years ago. At the time I think I was the only guy on the island who knew any BJJ. I'm glad I didn't mix it up with any Muay Thai fighters, although I was tempted :-)

@26 min: great discussion by one of the trainers about the fast evolution of MMA techniques and tactics. @32 min, the attraction of the fight game.

Hoxby: elite universities as venture investors in human capital

From a new paper by Stanford economist Caroline Hoxby. She doubts the utility of MOOCS for elite institutions within the model she advocates for their operations: as investors in human capital. See earlier posts Magical Mystery MOOCS , Selectivity of US colleges and returns to elite education.

More comments when I have some time ...

THE ECONOMICS OF ONLINE POSTSECONDARY EDUCATION: MOOCS, NONSELECTIVE EDUCATION, AND HIGHLY SELECTIVE EDUCATION

Highly Selective Postsecondary Education (HSPE)

In other studies, I model HSPE institutions as venture capitalists in advanced human capital because (i) they invest massively in each student whom they educate and (ii) they earn an equity-like return on their investments (Hoxby, 2012). That is, the HSPE institution itself covers the majority of the cost of a student's undergraduate education using donations from alumni. Even HSPE students who receive no financial aid pay for only some of the full cost of their education.

HSPE alumni donate a share of their perceived returns on the educational investment made in them, and most donations occur decades after graduation. Because HSPE alumni earn returns that exhibit a highly right-skewed distribution (in each classs, there are many alumni with solid professional careers but only a few Steve Jobs), it matters that donations are analogous to shares of returns and not to the repayment of a loan (the institution's investment plus some fixed rate of return).

It is crucial for understanding HSPE institutions' viability that they are modeled as investors, not as the sellers of a service. If an HSPE institution does not infuse value-added into students that is proportional to the institution's investment, alumni cannot pay the institution back later in life, no matter how much they wish to do so. Thus, at least on average, HSPE institutions' investments must earn returns or they will go bankrupt. Interestingly, the alumni who are the most generous donors are those who needed financial aid as students and who attribute their later success to the institution.6

Of course, this financing system only works if former students do in fact pay back the institution later in life. The methods by which HSPE institutions achieve this intergenerational virtuous circle are therefore crucial and discussed below. ...

“We basically have the whole genome sequence of an entire nation.”

Interesting comments from Kari Stefansson of deCODE (2012). See also this earlier video; Kong notes that multiple common variants can be used to tag rare variants ...

Bio-IT World: “We have fully sequenced now about 2,500 Icelanders, and publishing paper after paper about discoveries coming out of sequencing. As in everything, the proof is in the eating. We’re managing and mining these data effectively. There’s nothing miraculous about it, but it hasn’t constituted any destructive discovery in IT. We have reasonable software system to gather, store, and manage these data. We’re making a lot of discoveries out of them.”

“A lot has been written of late about the difficulty of managing manage all the data from whole genome sequencing. Analysts at big banks such as Goldman Sachs have been making bold statements about how important informatics inventions will be in biology because they will constitute the next disruptive discoveries coming out of biology, giving the impression that for the moment, this is an insurmountable difficulty in medicine. And that is just a crock of s—t!”

Stefansson says there are about 20 recently formed companies trying to develop software for genome sequencing interpretation, which he says deCODE has already achieved.

“We’re going from genotypes to phenotypes, not from a disease and looking for a variant. There are many companies trying to sell the scientific community methods to manage WGS data to make discoveries, and other companies trying to market software systems to manage clinical systems. And we have this all in place.”

Stefansson also says that deCODE plans to begin marketing later this year its system for managing sequence data, both for discovery and clinical sequencing – as well as release diagnostic tests in areas such as developmental disorders, prenatal testing, and cancer. Some of this may be done through spin-out companies.

deCODE’s principal sources of revenue comes from collaboration with big pharma, diagnostics tests, and scientific grants. ”Future growth will come from projects that will help integrate genetics into healthcare: software systems, diagnostic tests, and the way in which we help the pharma industry to use genetics.”

We have sequenced the whole genomes of 2,500 people. We have genotyped about 120,000 Icelanders with an Illumina chip. We can impute whole genome sequence down to variants with less than 0.1% frequency into about 370,000 Icelanders -- there are only 320,000 living today!”

“We basically have the whole genome sequence of an entire nation.”

deCODE has formed a collaboration with the National Hospital of Iceland to impute the genome sequence of all patients over the past 3 years (~300,000) and ask: in what indication, in what disease would WGS have the biggest impact on the selection of treatment and outcome? “Rather than doing it on a priori assumptions, we’ll do it by data mining,” he says. “Rather than doing it on a handful of people, we’ll do it on the basis of the whole nation. That shouldn’t take more than a year. We’re in a privileged position.”

So Money

An oral history of Swingers (Grantland).

Jon Favreau (Mike): When I set out to write Swingers, I didn’t know I was even writing a movie. My dad had given me a screenwriting program and I started the script just as an exercise to see if I could write a screenplay. Swingers is what came out.

Ron Livingston (Rob): Jon was really busted up over his breakup. That’s right about the age when your first long-term thing comes apart. Having your heart ripped out like that — that’s a lot different than the cheerleader who doesn’t like you back.

Favreau: I started writing, just drawing from the environment I was living in. I had characters loosely based on people I knew. None of the events were real; it was all a story that came out of my head without an outline.

Alex Désert (Charles): I like to say Swingers was us times 10. I wish I could be that cool.

...

Livingston: Vince and I, and a couple of other people — Alex Désert, Ahmed Ahmed, who’s in the movie — we started doing all of these staged readings for potential buyers of this script. They came in all shapes and sizes. Every three months or so, we’d get together in somebody’s living room and rehearse for a while and then go to some empty theater space and do it for some guy who had Saudi parking lot money. ...

Vaughn: The reading would always play phenomenally. We did this for over a year and would get huge laughs, great responses. But the business model was always a problem. You have a bunch of guys that don’t really mean anything to Hollywood. Jon had done more than the rest of us, but wasn’t a big enough name to open a movie. ...

Wurmfeld: Literally every single word that comes out of Vince’s mouth is on the page. That’s what totally blows me away about Jon’s writing — his ability to get someone’s voice, because I think that’s not an easy task. One might think that Vince is improvising, and certainly he can, but I just was amazed that all those jokes and stuff were actually on the page.

Livingston: He grabbed “You’re so money” from the Spike Lee–Michael Jordan commercials, where Spike Lee called Michael Jordan “Money,” you know, “Like the shoes, Money.” Nobody was really doing that, I think, other than just Spike Lee and Michael Jordan. So when the movie came out, that was still kind of a new thing.

STEM pays

Peak annual earnings for engineering graduates were even higher at $98k. Source: How Liberal Arts and Sciences Majors Fare in Employment: A Report on Earnings and Long-Term Career Paths.

When I was a new faculty member at Oregon I was told with great earnestness by our President (a JD with limited understanding of STEM or serious academic research) that liberal arts majors made more than science and engineering graduates in the long run. This encounter catalyzed my suspicion that those who run the world often don't understand how it actually works.

Better figures here.

Compressed sensing and genomes v2

We placed a new version of our compressed sensing and genomics paper on arxiv. For earlier discussion see here and here. New results concern the effect of LD (linkage disequilibrium) on the method, and a demonstration that choosing a large L1 penalization parameter allows the selection of loci of nonzero effect at sample sizes too small to recover the entire effects vector -- see below, including figure.

As I've mentioned before, one practical implication of this work is that of order 30 times sparsity s (s = total number of loci of nonzero effect) is sufficient sample size at h2 = 0.5 to extract all the loci. For a trait with s = 10k, somewhat less than a million individuals should be enough.

Application of compressed sensing to genome wide association studies and genomic selection (http://arxiv.org/abs/1310.2264)

We show that the signal-processing paradigm known as compressed sensing (CS) is applicable to genome-wide association studies (GWAS) and genomic selection (GS). The aim of GWAS is to isolate trait-associated loci, whereas GS attempts to predict the phenotypic values of new individuals on the basis of training data. CS addresses a problem common to both endeavors, namely that the number of genotyped markers often greatly exceeds the sample size. We show using CS methods and theory that all loci of nonzero effect can be identified (selected) using an efficient algorithm, provided that they are sufficiently few in number (sparse) relative to sample size. For heritability h2 = 1, there is a sharp phase transition to complete selection as the sample size is increased. For heritability values less than one, complete selection can still occur although the transition is smoothed. The transition boundary is only weakly dependent on the total number of genotyped markers. The crossing of a transition boundary provides an objective means to determine when true effects are being recovered; we discuss practical methods for detecting the boundary. For h2 = 0.5, we find that a sample size that is thirty times the number of nonzero loci is sufficient for good recovery.

An excerpt from the new version and corresponding figure:

... CS theory does not provide performance guarantees in the presence of arbitrary correlations (LD) among predictor variables. However, according to our simulations using all genotyped SNPs on chromosome 22, L1-penalized regression does select SNPs in close proximity to true nonzeros. The difficulty of fine-mapping an association signal to the actual causal variant is a limitation shared by all statistical gene-mapping approaches—including marginal regression as implemented in standard GWAS—and thus should not be interpreted as a drawback of L1 methods.

We found that a sample size of 12,464 was not sufficient to achieve full recovery of the nonzeros with respect to height. The penalization parameter λ, however, is set by CS theory so as to minimize the NE. In some situations it might be desirable to tolerate a relatively large NE in order to achieve precise but incomplete recovery (few false positives, many false negatives). By setting λ to a strict value appropriate for a low-heritability trait, we found that a phase transition to good recovery can be achieved with smaller sample sizes, at the cost of selecting a smaller number of markers and hence suffering many false negatives. ...


Figure 7 Map of SNPs associated with height, as identified by the GIANT Consortium meta-analysis, L1-penalized regression, and standard GWAS. Base-pair distance is given by angle, and chromosome endpoints are demarcated by dotted lines. Starting from 3 o’clock and going counterclockwise, the map sweeps through the chromosomes in numerical order. As a scale reference, the first sector represents chromosome 1 and is ∼ 250 million base-pairs. The blue segments correspond to height-associated SNPs discovered by GIANT. Note that some of these may overlap. The yellow segments represent L1-selected SNPs that fell within 500 kb of a (blue) GIANT-identified nonzero; these met our criterion for being declared true positives. The red segments represent L1-selected SNPs that did not fall within 500 kb of a GIANT-identified nonzero. Note that some yellow and red segments overlap given this figure’s resolution. There are in total 20 yellow/red segments, representing L1-selected SNPs found using all 12,454 subjects. The white dots represent the locations of SNPs selected by MR at a P-value threshold of 10−8. (Click for larger version!)

Making Oblivion

I just caught Oblivion on cable -- it's visually stunning. The Skytower or glass house was shot in a very interesting way, as shown in this video. If you like modern design and architecture you'll love this.

Keep your head down, and smile

In 2005 I posted this interview of Lee Kuan Yew that appeared in Der Spiegel. It seems to me that the PRC government may not have the patience to bide its time as recommended by Lee. Once again, he appears prophetic.

SPIEGEL: The Chinese Government is promoting the peaceful rise of China. Do you believe them?

Mr. Lee: Yes, I do, with one reservation. I think they have calculated that they need 30 to 40 -- maybe 50 years of peace and quiet to catch up, to build up their system, change it from the communist system to the market system. They must avoid the mistakes made by Germany and Japan. Their competition for power, influence and resources led in the last century to two terrible wars.

SPIEGEL: What should the Chinese do differently?

Mr. Lee: They will trade, they will not demand, "This is my sphere of influence, you keep out". America goes to South America and they also go to South America. Brazil has now put aside an area as big as the state of Massachusetts to grow soya beans for China. They are going to Sudan and Venezuela for oil because the Venezuelan President doesn't like America. They are going to Iran for oil and gas. So, they are not asking for a military contest for power, but for an economic competition.

SPIEGEL: But would anybody take them really seriously without military power?

Mr. Lee: About eight years ago, I met Liu Huaqing, the man who built the Chinese Navy. Mao personally sent him to Leningrad to learn to build ships. I said to him, "The Russians made very rough, crude weapons". He replied, "You are wrong. They made first-class weapons, equal to the Americans." The Russian mistake was that they put so much into military expenditure and so little into civilian technology. So their economy collapsed. I believe the Chinese leadership have learnt: If you compete with America in armaments, you will lose. You will bankrupt yourself. So, avoid it, keep your head down, and smile, for 40 or 50 years.

SPIEGEL: What are your reservations?

Mr. Lee: I don't know whether the next generation will stay on this course. After 15 or 20 years they may feel their muscles are very powerful. We know the mind of the leaders but the mood of the people on the ground is another matter. Because there's no more communist ideology to hold the people together, the ground is now galvanised by Chinese patriotism and nationalism. Look at the anti-Japanese demonstrations. ...

For more on this topic, I recommend this talk by Lieutenant General Noboru Yamaguchi.

CHINA’S RISE, AMERICA’S PIVOT, AND JAPAN’S CHOICE: The security landscape in the Asia-Pacific region will mainly be determined in the long term by two significant actors, the People’s Republic of China and the United States. China’s rise has been and will be an extremely prominent factor; while how fast and in which direction China proceeds are open questions. The US has taken a significant step to stay involved in the region by declaring that it will rebalance towards the Asia-Pacific. For architects of Japan’s security strategy, it is crucially important, according to Lt Gen Yamaguchi, to, first, scrutinise both the negative and positive impacts of China’s rise without any bias; second, have a clear picture of the rebalancing by the US with an understanding of the roles of American allies for proper implementation of this policy; and third, develop courses of action and possible policies for Japan to better serve the region’s peace and stability.

The lecture has been made possible by the support of Japan's Ministry of Foreign Affairs.

Flying robot jellyfish

Livescience: A new teeny-tiny robot flies through the air like a jellyfish swims. The jellyfish flier is a strange sight — it looks a little bit like a Chinese lantern that's developed a hankering for the open skies — but its unique design keeps it from tipping over without the use of sensors or external controls. 

That talent could make it handy for maneuvering in small spaces, said its inventor Leif Ristroph, a postdoctoral researcher at New York University. "What's cool is you can actually build these flying things yourself," Ristroph told LiveScience. "All the components I used to make this, they cost about $15 and they're available on hobby airplane websites." 

Ristroph, whose background is in physics, studied insect flight for his doctorate. Insects are intriguing to physicists and engineers, because of their flying abilities. Contrary to what might be expected, it's harder to engineer something tiny to fly than something big, Ristroph said. ...

Cancer genomics

A great article in Esquire on the frontier of cancer genomics. It's only a matter of time before many wealthy individuals diagnosed with cancer seek out this standard of care. (I'm told that sequencing alone can lead to actionable conclusions in more than a few percent of all cases, even now.) The resulting advances will hopefully eventually reach everyone else.

Biomathematician Eric Schadt, shown above.

Esquire: ... When [Schadt] graduated high school, he joined the Air Force with the idea of subjecting himself to the rigors of Special Forces training. Instead, he blew out his shoulder on a climb, and the Air Force tried to salvage its investment by putting him through a battery of tests. He took them; when the scores came back, he was asked by stunned superiors if math had always come easily to him. Then he was sent to college and undertook the task of complete intellectual self-transformation. He received an undergraduate degree in applied mathematics and computer science at Cal Poly and his master's in pure mathematics at UC Davis. Pure math was, to him, the Special Forces of the mind—he took it because it was so hard, and he wanted to find out just how smart he was. He was pretty smart, as it turned out, but he despaired of working on problems that existed on the level of pure abstraction and had no bearing on the problems of the world. It seemed like, well, a sin. He went to UCLA to get a Ph.D. in the emerging field of biomathematics. The one problem was that the degree required a Ph.D.-level mastery of molecular biology, and the last biology course he'd taken was in high school. So he taught himself by reading textbooks. It wasn't hard. Pure math was hard. Molecular biology, after pure math, struck him as ridiculously easy.

... Schadt told Merck that this was a strategy doomed to fail, because disease arose not from single genes or pathways but rather out of vast networks of genes and pathways whose interactions could be understood only by supercomputers guided by abstruse algorithms. Evangelical still, though now evangelical on behalf of irreducible complexity, he asked Merck to remake itself in the image of the network model he was determined to pioneer. Merck declined and Schadt headed to Silicon Valley, to the land of data.

... In September, Mount Sinai announced that he would be head of the newly created Institute for Genomics and Multiscale Biology. A little more than a year later, Sinai announced that Schadt's operation would be renamed the Icahn Institute, just as the entire medical school would be renamed the Icahn School of Medicine at Mount Sinai. For the privilege, Carl Icahn had given Eric Schadt $150 million to claim the future of biology.

The article focuses on the treatment of Stephanie Lee's colon cancer.

Esquire: ... The first thing that needs to be understood about Stephanie's data is that there would be a lot of it. From the samples of Stephanie's blood, the gene sequencers would generate the data about the genes in her "germline"—the genes (and the gene mutations) that she inherited from her parents and that existed in every cell of her body. From the samples sliced from her colon, the sequencers would generate data about her cancer and the mutations that produced it. But the data would be raw. It would contain millions of bits of genetic information, each one a sentence in the horror story that Stephanie's cancer was telling—and all those sentences, at least initially, adding up to a bewildering Babel. The data would exist right on the edge of incoherence; then Schadt and his scientists would strive both to make sense of it and complicate it. That's their trademark, and why they need a supercomputer. The genes that Stephanie was born with would be compared with the genes that were driving Stephanie's cancer. The genes that were driving Stephanie's cancer would be compared with the vast libraries of reference data-bases that already exist on all kinds of cancers. Then they would be plotted against the "network models" that the Icahn Institute is constructing, the millions of individual data points mined for their billions and even trillions of connections.

... Schadt's scientists—his biologists and his mathematicians—were from all over the world. Many had followed Schadt from the West Coast, but before they came to America, they lived in China and India and Russia. Now they had access to near infinities of information; indeed, they would soon have access to the near infinity of information generated by the DNA of a woman in Mississippi who had been given no information at all, except the information that she was going to die. It was a point lost on no one, least of all Schadt.

... A month earlier, Cagan had started doing something that he said "had never been done before." He started creating "personalized flies" for cancer patients. He took the mutations that scientists like Schadt had revealed and loaded them into flies, essentially giving the flies the same cancer that the patient had. Then he treated them. "Why a fly? You can do this in a fly. You can capture the complexities of the tumor."

... By October 11, however, Cagan already had "one possible drug suggestion for her"—or one possible combination of drugs, since he always tests at least two at a time.

... Now the oncologists at Mount Sinai were asking Schadt and Cagan if there was something they could do in their own intractable cases. And Dr. Holcombe was asking what studies had to be done to incorporate personalized therapy into the standard of care…a standard of care derived in part from none other than little ol' Stephanie Lee. There had been the imposition of obstacles at every step of the way, and the odds against her remaining on what Schadt called "a path to fight on" had beggared even his mathematical imagination. And yet here she was. Her tumor had become, in Ross Cagan's words, "at this point in time, the most fussed-over tumor that I'm aware of." ...

Here's a peek at some of the computing infrastructure behind this kind of work.

US economic mobility data

2017 Update: the Pew report has moved; old links below are broken.

A commenter points me to the Pew Trust study on US economic mobility. I excerpt some interesting figures from the report below, but it is worth reading in full. Much of the data comes from NLSY. Note that mobility data over quintiles does not address the increasing concentration of wealth at the top: see US wealth inequality and Inside the 1%.

Americans born in middle income quintile families have roughly equal probability of ending up in any of the other quintiles. Being born in a top quintile family is far from a guarantee of persistence in the top quintile, and similarly for the bottom quintile.

White men have slightly more upward mobilility than other groups.

The outcome distribution for top quintile (family background) black men is most similar to that for median quintile white men.

Superior cognitive ability is strongly indicative of upward mobility from the lowest quintile. This is true for both blacks and whites. Individuals who are cognitively at the median in the overall population have an 80% chance of moving up and out of the bottom quintile.

Pew: Individuals with higher test scores in adolescence are more likely to move out of the bottom quintile, and test scores can explain virtually the entire black-white mobility gap. Figure 13 plots the transition rates against percentiles of the AFQT test score distribution. The upward-sloping lines indicate that, as might be expected, individuals with higher test scores are much more likely to leave the bottom income quintile. For example, for whites, moving from the first percentile of the AFQT distribution to the median roughly doubles the likelihood from 42 percent to 81 percent. The comparable increase for blacks is even more dramatic, rising from 33 percent to 78 percent. Perhaps the most stunning finding is that once one accounts for the AFQT score, the entire racial gap in mobility is eliminated for a broad portion of the distribution. At the very bottom and in the top half of the distribution a small gap remains, but it is not statistically significant. The differences in the top half of the AFQT distribution are particularly misleading because there are very few blacks in the NLSY with AFQT scores this high.

There is a tide

This novel reminds of recent dystopian movies and books such as Elysium, The Road, Never Let Me Go and Cloud Atlas. Our zeitgeist recognizes the ever accelerating gap between haves and have-nots, and these novels and films reflect that collective consciousness. The gap always existed between the rich and poor worlds, but now it exists within developed countries like the US, threatening broad swathes of the former middle class.

Leonard Lopate Show: Chang-Rae Lee talks about his new novel, On Such a Full Sea, set in a future, when a long-declining America is strictly stratified by class. Abandoned urban neighborhoods have become high-walled, self-contained labor colonies. The members of the labor class work to provide quality produce and fish to elite villages. In this world lives Fan, a female fish-tank diver, who leaves her home in the B-Mor settlement (once known as Baltimore), when the man she loves mysteriously disappears.

Interestingly, the wealthy dwellers of the Charters allow the top 1 percent of children from the labor colonies to ascend in class. The main character Fan's brother was adopted by a Charter family, and she seeks him out for help. The genomic angle manifests in the disappearance of Fan's love, who is dispatched (for study?) because of a rare mutation that makes him resistant to the still-deadly "C-diseases" (cancer?).

An excerpt from the book.

... When our ancestors were first brought here—the archival vids and pix show them rolling in on fleets of shiny silver company buses—the air was to them fresh and clear, just like in the image of the roofless row house, and when they stepped out, they must have been entranced by the scant briny notes of the harbor waters, breathing them in deep. And think of how startled they might have been by the strange brand of tidiness in this place (once known as Baltimore) and other abandoned cities that settlers were sent to in other eastern and midwestern states, this preservation by dint of absence, such that after they gathered their luggage from the curb and were shuttled by carts to the houses assigned to them, our and your and Fan's forebears among them, their gasps were not of trepidation or disappointment but of gratitude and relief.

Indeed, it's difficult for us to understand how genuinely grateful they were; we glance around B-Mor now and it's impossible to imagine how our people could have felt that way (how time and safety and a filled belly rapidly evolve us!), to be presented with so depleted a cityscape and still have a heart-surge of excitement. The legendary Wen Shurbao, who would be our first and only mayor, reportedly exhorted his brethren by invoking the classic proverb: "Our generation will plant the trees. The next will enjoy the shade."

Surely there were the discontented among those originals, but which of them could deny the promise of this place? Here was an entire community, ready for revitalization. Yes, the houses were basically shells, but in fact many still had roofs and walls and sturdy stairs; yes, few had any boilers, but the majority had salvageable wiring and plumbing; yes, the floors had to be scraped and sanded and refinished, every cabinet and counter scrubbed and disinfected of the leavings of birds and vermin and insects, and yet what activity offers more immediate, honest gratification than shining up a seemingly ruined surface back to the distinctive grain of its essence?

... We should concede that unlike the experience of most immigrants, there was very little to encounter by way of an indigenous population. There were smatterings of them, to be sure, pockets of residents on the outskirts of what is now the heart of B-Mor, these descendants of nineteenth-century African slaves and twentieth-century laborers from Central America and even bands of twenty-first-century urban-nostalgics, all of whom settled the intimate grid of these blocks and thrived for a time and, for reasons that history can confidently trace and identify but never quite seem to solve, inexorably declined and finally disappeared. Our predecessors had the unique advantage of being husbanded by one of the federated companies, rather than the revolving cast of governmental bodies that overreached in their efforts or were disastrously neglectful, all of them downright clueless. The originals were brought in en masse for a strict purpose but with their work- and family-centric culture intact, such that they would not only endure and eventually profit the seed investors but also prosper in a manner that would be perpetually regenerative.

... Perhaps it was the same with our originals, though in a different circumstance. They went about their first labors, renovating the row houses in the same way, it turns out, that certain antique American communities used to do, the foreman or forewoman of each block marshaling all its residents to converge on one address and revamp, say, the bathrooms or kitchen, the museum clips just like a science class vid of hundreds of ants tugging a sourball-sized rock. You can picture it now. They'd go from one house to the next, right on down the block, this mobile, instantly adaptive assembly line, each person assigned a function, with the children passing beach pails of dust and rubble in a brigade, the elderly offering sips of cool chrysanthemum tea from canteens, even the unwell propped up in chairs close by or even inside the site, so that they might lend moral support or learn by watching. ...

Julius Caesar, Act 4, Scene 3,

BRUTUS
Under your pardon. You must note beside,
That we have tried the utmost of our friends,
Our legions are brim-full, our cause is ripe.
The enemy increaseth every day.
We, at the height, are ready to decline.
There is a tide in the affairs of men,
Which, taken at the flood, leads on to fortune;
Omitted, all the voyage of their life
Is bound in shallows and in miseries.
On such a full sea are we now afloat,
And we must take the current when it serves
Or lose our ventures.

Meritocrats

An essay by the late historian Tony Judt, composed as he slowly died of motor neuron disease.

Meritocrats

I came up to King’s College, Cambridge, in 1966...

My King’s was the very incarnation of meritocratic postwar Britain. Most of us got where we were by doing well in exams and, to a striking extent, we pursued occupations that reflected our early talents and interests. The cohort of Kingsmen who came up in 1966 stand out in their choice of careers: more than any group before or since, we opted for education, public service, the higher reaches of journalism, the arts, and the unprofitable end of the liberal professions.

It is thus altogether appropriate that the most promising economist of our generation—Mervyn King—should have ended up as the governor of the Bank of England, rather than an investment banker or hedge funder. Before our time, talented Kingsmen doubtless followed similar paths. But a glance at the obituaries of an older generation reveals just how many of them returned to the family business or to the traditional professions of their fathers and grandfathers.

As for those who came after, it is depressing to record how quickly and in what numbers the graduates of the 1970s and since resorted to the world of private banking, commerce, and the more remunerative reaches of the law. Perhaps one should not blame them; in our time, jobs were still plentiful and we could bask in the waning rays of postwar prosperity. All the same, it’s very clear that our elective affinities lay elsewhere.

... I served for a while as a very junior member on the College Fellowship Electors with Williams, John Dunn, Sydney Brenner (the Nobel Prize winner in medicine), Sir Frank Kermode, Geoffrey Lloyd (the historian of ancient science), and Sir Martin Rees (the Astronomer Royal). I have never lost the sense that this was learning: wit, range, and above all the ability (as Forster put it in another context) to connect. 

My greatest debt, though I did not fully appreciate it at the time, was to Dunn, then a very young college Research Fellow, now a distinguished professor emeritus. It was John who—in the course of one extended conversation on the political thought of John Locke—broke through my well-armored adolescent Marxism and first introduced me to the challenges of intellectual history. He managed this by the simple device of listening very intently to everything I said, taking it with extraordinary seriousness on its own terms, and then picking it gently and firmly apart in a way that I could both accept and respect.

That is teaching. It is also a certain sort of liberalism: the kind that engages in good faith with dissenting (or simply mistaken) opinions across a broad political spectrum. No doubt such tolerant intellectual breadth was not confined to King’s. But listening to friends and contemporaries describe their experiences elsewhere, I sometimes wonder. Lecturers in other establishments often sounded disengaged and busy, or else professionally self-absorbed in the manner of American academic departments at their least impressive.

... For forty years, British education has been subjected to a catastrophic sequence of “reforms” aimed at curbing its elitist inheritance and institutionalizing “equality.” The havoc wrought in higher education was well summarized by Anthony Grafton in this magazine, but the worst damage has been at the secondary level. Intent upon destroying the selective state schools that afforded my generation a first-rate education at public expense, politicians have foisted upon the state sector a system of enforced downward uniformity.

... Universities are elitist: they are about selecting the most able cohort of a generation and educating them to their ability—breaking open the elite and making it consistently anew. Equality of opportunity and equality of outcome are not the same thing. A society divided by wealth and inheritance cannot redress this injustice by camouflaging it in educational institutions—by denying distinctions of ability or by restricting selective opportunity—while favoring a steadily widening income gap in the name of the free market. This is mere cant and hypocrisy. ...

See also Defining Merit and Les Grandes Ecoles Chinoises.

There and back again

A day at the beach.

Obligatory selfie.

The kids caught two lizards in the backyard -- "Larry" and "Barry" :-)

Rose Bowl victory for MSU.

The scene at Detroit airport  :-(

I have a lot more photos but the wife has decreed that family photos are not to appear online.

PS I would have preferred to stay at home and work but other forces prevailed.

The Gene Factory (New Yorker)

LETTER FROM SHENZHEN: THE GENE FACTORY

A Chinese firm’s bid to crack hunger, illness, evolution—and the genetics of human intelligence.

BY MICHAEL SPECTER

A well-balanced article about BGI in the New Yorker.

I am misquoted about our current ability to predict height from genomic information, despite an hour on the phone with the Harvard-educated fact checker. There is also some confusion in that paragraph concerning the correlation between height and IQ and its relevance to the broader discussion  :-(

Chris Chang has the wise, last words in the article :-)  Click for larger version.

Research rankings

Some notes for my own use on research rankings by discipline for Michigan State University. The ranges are 95% confidence intervals using the R-rank (including, among other things, publications, citations, grants and awards) from the 2010 NRC evaluation (see, e.g., here). Rank ordered by, e.g., central value of R range, most of the departments below come out in the top 20-30 in the US. More data: S rankings in various fields.

Physics                                     17-41
Chemistry                                 21-59
Computer Science                    12-32
Math                                         11-43
EE/CE                                      47-87
Biochem/MoBio                      54-96
Neuroscience                            72-89
Plant Biology                            6-32

In the NSF HERD (total R&D expenditures) data MSU is ranked 36th in the US, with 2012 expenditures just over $500 million. If research hospital numbers are excluded (MSU has historically had a teaching medical school, although this is changing), the MSU rank would be quite a bit higher.