May 30, 2003
a break

Tomorrow morning I'm off for Kauai and Maui for 8 days. It will be a welcome break. Normally before I visit someplace I "study up" on it--learn its culture, its history, even its geography. I've been so busy the last couple of weeks (in spite of having graduated) that I haven't had a chance. I'll be cramming on the airplane.

Today I did an interview with WTVD which will air sometime before the Race for the Cure on 14 June, and also did a group interview over speakerphone with a reporter from the Wall Street Journal. After the drill of last week, talking on camera really isn't bad at all. Basically, the key is to ignore it completely and just address your interviewer. And like most news reporters, she had the smoothness, and drawn-out, comforting intonation that makes having a conversation in spite of lights and camera easy.

I guess I'm a media slut to some extent because I really don't mind doing this stuff at all.

Still, it will be nice to stretch out on a sandy, Pacific beach watching the cerulean waves roll in, reading the American College of Surgeons' Advanced Trauma Life Support, 1997 Edition.

Posted by erich at 04:54 PM
May 28, 2003
herding cats

I'm getting the impression that getting an industry-academic collaboration to really work is like herding cats. Inherently, everyone moves in different directions. The one thing everyone has in common is that they want to get the maximal return on minimal investment--it's just the returns and investments are different for different parties. If our breast cancer project were solely a commercial enterprise, the logistics of setting it up with be so much easier because the ultimate goal would be to deliver a product that satisfies a demand and returns a profit; everything would fall in line behind that.

But what we're trying to do is something different.

We want to develop interactions that provide a return of course, but not just in $$$, but in terms of science, and the infrastructure that can translate good science into good products that will benefit patients in a real way.

With the difficulties getting everyone moving in the same direction, this may as well be alchemy. On the other hand, you can see this as a challenge, a puzzle to figure out, which depends as much on cleverness as on being able to cajole and wheedle. As long as there isn't too much hubris involved, I suppose it's doable. I truly hope this isn't a quixotic enterprise.

But having a good lawyer helps.

Posted by erich at 01:56 PM
May 26, 2003

Thanks to my friend Ron Eng, who is partnered with my sister in their little firm Formactiv, the house was remodeled (painfully, as all remodeling projects are wont to be) last year. In exchange for the angst of being thrown out of my own house and having workpeople tramping about, I now have this space to work in.

Posted by erich at 12:36 PM
May 25, 2003

Posted by erich at 03:35 PM
May 24, 2003
contrasts & legs

Posted by erich at 05:05 PM
what starbucks cannot supply

Columbian Mesa de Los Santos ed this morning just past 2nd crack.

Posted by erich at 12:21 PM
rescue of the shuttle crew was "conceivable"

A classmate's father, who recently retired from managing payloads for the Shuttle and worked for Grumman in the glory days when they built the LEM, sent me an application for NASA's Astronaut Corps a few years ago. I kept it, thinking that once I graduated, I might just send it in.

With that application in hand, I have thought very seriously at times of what it might mean to send it in. I spent a great deal of time (and sacrificed many trees) printing out public reports and feasibility studies on a manned Mars mission, and wondered what it might be like being a mission physician on a two year sojourn on Mars and in space. I've read Gene Krantz's autobiography, and Andrew Chaikin's excellent history of the Apollo program. No doubt much of this was daydreaming, and now I have many earthbound obligations, but the Space Program represented for me a combination of both the most visceral of human yearnings, exploration, and the idealism that makes such an urge material, whether in the form of a wooden bark, or a Saturn V.

It's discouraging that NASA is a very different institution now. The blame doesn't reside in NASA alone. Society's priorities have changed. Most respected historians of the Space Program note that the triumphant arrival of American astronauts on the Moon was a sublimation of the arms race with the Soviet Union, and without that ideological/military competition, the urge to pour a civilization's wealth and talent into such a focused goal has dissipated. And with this dissipation comes bureaucratization and diminution of ambition. And worse, the pernicious fatalism that sacrificed seven astronauts.

Perhaps the greatest indictment of this institutional passivity is Adm. Harold Gehman's statement yesterday that, supposing the danger to the Shuttle's crew had been seen with the proper urgency, the Atlantis, which was being readied for the next mission, could have been sent up on a rescue mission with an abbreviated pre-flight.

Drawing a comparison with military pilots, Gehman suggested that there is an "unwritten contract" that the utmost will be done to rescue stranded crew. No doubt this rescue mission would have been exceedingly difficult--there is no provision for two Shuttles to dock to one another, so astronauts would have had to space walk from the Columbia to the Atlantis--and risky for both crews, but the fact that there was no opportunity to even entertain the notion of a rescue is tragic.

Apollo 13 represents the best of American values, absolute refusal to accept failure and the ingenuity of hundreds of engineers and controllers to bring three men in a dead spaceship back home. What happened to this spirit? It is still hard for me to digest the notion that individual engineers can have identified danger to the crew, but that the Shuttle Program's culture accepted that nothing might be done, and fatalistically let the crew perish in a fireball above Texas.

Posted by erich at 11:10 AM
May 23, 2003
SARS, a mongoose virus?

Articles in the Washington Post and the NY Times suggest that there's the possibility that the coronavirus implicated in the SARS epidemic may have originated in exotic wild game that are served as delicacies in Asia. Researchers from China and the University of Hong Kong went to a live animal market in Guangdong (Canton) province and tested 25 animals there. Among them were six the Masked Palm Civets (an animal related to the mongoose)

which demonstrated a coronavirus almost identical to the SARS example in both their feces and saliva. A Raccoon Dog was found to have "genetic evidence" of the coronavirus in its feces (probably a PCR test), and a Chinese Ferret Badger possessed antibodies to the coronavirus in its blood (in other words, the virus itself wasn't found, but antibodies to defend against the virus were).

This evidence is highly suggestive. The Washington Post article has more scientific details than the Times piece. It reports that sequence of the coronavirus from these civets was identical to the human SARS virus save for 29 basepairs.

It is already known that the human version possesses 29.7-thousand letters in its genetic code. The virus found in the animals differs by just 1/1000th of this genetic code.

If the virus did indeed "jump" from an animal reservoir to human, the mechanism may be that the virus lost a small bit of genetic code that truncate a peptide sequence as small as nine amino acids in the virus's external shell. Somehow this modification made this virus highly pathogenic in humans.

This is all hypothesis right now, there are lots of questions to be answered. It's definitely possible that these animals were infected by yet another host that is the "true reservoir". It would also be interesting to come up with a mechanism of how this difference in 29 letters of genetic code developed.

Eating exotic wild game has a long and deeply entrenched tradition in Chinese culture. Masked palm civets are traditionally eaten during the autumn and winter to "help people withstand the winter". It's more likely that, supposing the virus made a species jump, that it occurred during the preparation of food, i.e. slaughtering and cleaning the animal, rather than eating the animal (where cooking would have inactivated the virus).

Posted by erich at 06:02 PM
one week

To this...

Posted by erich at 01:19 PM
May 21, 2003
sound bites

I had an interesting morning.

The news office wanted to video me doing sound bites for our individualized breast cancer prognosis paper. To distill abstruse scientific concepts into digestible nuggets, you have to think about 10 seconds ahead of what's coming out of your mouth. I don't know how Ari Fleischer does it every day.

I confess it was sort of fun...a challenge trying to assemble my stream of consciousness into something coherent. But by an hour and a half, not so coherent.

Posted by erich at 09:15 PM
May 20, 2003
reportage & complexity

I just had a call from a woman from the New Yorker who's putting together content for an "advertorial" supplement on breast cancer. Nothing like finding myself in limbo land between ads for drug companies and editorial content--even if it's in the New Yorker.

She was talking/thinking/typing at the same time (in fact, she even forgot to introduce herself until I asked her a couple of minutes later). This makes me uneasy because misstatements are easy to make in science reporting. It's not that we should subject the general public to the stultifying prose that you see in a scientific journal, but I think science writers have to be good filters; and being a good filter entails getting the complexity of the full scientific story first and then simplifying. If you simplify right off the bat, you're going to get something wrong. I think I'm pretty good at explaining complex concepts, providing that I'm given time to explain them, but the reporters I've talked to always seem to be in a rush.

Hey, there are patients out there, we want to get things right.

Anyway, she wanted me to say how many genes we require for our predictive models. I tried to explain that we shouldn't pin ourselves to a small subset of genes because breast cancer is complex, and different features of the tumor's behavior are going to involve different groups of genes. She retorted that NKI/Bernards/Friend got down to 70 genes, and that Chuck Perou got down to a couple of hundred.

I guarantee that 70 genes will never tell the whole story for breast cancer.

Posted by erich at 07:28 PM
sciencentral on both papers

This piece addresses personalized medicine for cancer in general, and covers the CODEx Project along with the NKI/Bernards/Friend group and their current clinical trials.

Posted by erich at 12:02 PM
a little press

The Atlanta Journal-Constitution has a small piece (via wire service) on the latest Nature Genetics paper. Actually a bit surprising considering this paper is pure science and not as overtly clinical as the other.

Posted by erich at 09:47 AM
May 19, 2003

This weekend I decided to turn off my cellphone. I have a blissful 5 weeks before my pager will be vibrating out of its holster by the minute, before I'll be the scut monkey for attendings and senior residents alike, before I wear white pants seven days a week, so I thought it would be nice to reduce the electronic interruptions of my life for at least a short while.

Not bad.

I just checked my messages. A lady from Asbury Park, NJ had seen a newspaper article about the Lancet paper, had somehow gotten my cellphone number, and wanted to know if her daughter could somehow take the test. I'll call her back tomorrow morning, but I don't know what to do if I start getting many such calls. I don't think I can personally answer all of these. And how did she get my cellphone number?

...Well I just called her, nice middle-aged lady just trying to help out her daughter. I think she was confusing our assay with the BRCA1/2 test available through Myriad Genetics for women at high-risk for familial breast cancers. I like talking to patients and their families, so fielding this one call wasn't bad.

Forgot to ask her if she knew Bruce Springsteen.

I must talk to people in lab about handing out my mobile number.

Posted by erich at 08:09 PM
devant garde

The New York Times Magazine has a long piece about the revival of architecture since Gehry completed the Bilbao Guggenheim. The writer, Arthur Lubow, calls this the "Bilbao Effect", arguing that Bilbao proved that showcase architecture can convert a rusty backwater shipbuilding town to a place people go out of their way to visit. Another example he cites is Daniel Liebeskind's Holocaust Museum in Berlin, which hundreds of thousands visited daily even before exhibits were in place.

All of a sudden, "unbuildable" architects like Zaha Hadid, Bernard Tschumi, and Rem Koolhaas seem to be getting commissions.

This is very hopeful, but consider this: in New York, a city that considers itself a World cultural capital, if not the World's cultural capital, the only recent building of any consequence is Christian Portzamparc's LVHM Headquarters. I suppose one could point to Koolhaas's Prada store, but out of retail architecture are not monuments made. In fact, the most monumental (at least in scale and popularity) new architecture in New York is the mainstream retail fantasy of Times Square.

Yes, architecture has's just that most of it is in Europe or Asia.

OK, it's true that Gehry has a good number of commissions in the US--most recently the Richard B. Fisher Center for the Performing Arts at Bard College--but what I'm saying is that interesting and innovative design still hasn't inserted itself in the common cultural discourse.

It still surprises me when I meet some hipster with trendy glasses and a Helmut Lang-ish wardrobe who can't tell the difference between Eames and Ethan Allen. There's a disjunction here that I don't understand. I guess the closest we've gotten to design ubiquity is the Aeron chair; now more an icon of the bust than an harbinger for good design.

Almost 20 years ago many hoped that Maya Lin's Vietnam Memorial might wake the country out of its architectural lassitude, then we wimped out and built an utterly conventional and figurative statue of three soldiers right next to it. In spite of this I'm watching what's happening with Ground Zero with a measure of hope.

Maybe I'm depressed because my favorite local design store went out of business...

Posted by erich at 04:35 PM
May 18, 2003
nature genetics paper published

This sure has been a week for papers. On the heels of last week's paper published in The Lancet, my paper "Gene expression phenotypic models that predict the activity of oncogenic pathways" is released for online publication today. Here's the link to Nature Genetics's Advance Online Publication.

Essentially, I manipulated cells to turn on genes implicated in cancer, "interrogated" these cells for genome-scale expression data, and used this data to develop robust predictive models for the activity of these cancer genes, also known as "oncogenes". I tested the models in normal synchronized proliferating cells (virtually all oncogenes have normal functions in regulating cell growth--they become cancer genes when they somehow become deranged), and in tumor tissues, with good success.

High-density DNA microarrays measure expression of large numbers of genes in one assay. The ability to find underlying structure in complex gene expression data sets and rigorously test association of that structure with biological conditions is essential to developing multi-faceted views of the gene activity that defines cellular phenotype. We sought to connect features of gene expression data with biological hypotheses by integrating 'metagene' patterns from DNA microarray experiments in the characterization and prediction of oncogenic phenotypes. We applied these techniques to the analysis of regulatory pathways controlled by the genes HRAS (Harvey rat sarcoma viral oncogene homolog), MYC (myelocytomatosis viral oncogene homolog) and E2F1 ,E2F2 and E2F3 (encoding E2F transcription factors 1, 2 and 3, respectively). The phenotypic models accurately predict the activity of these pathways in the context of normal cell proliferation. Moreover, the metagene models trained with gene expression patterns evoked by ectopic production of Myc or Ras proteins in primary tissue culture cells properly predict the activity of in vivo tumor models that result from deregulation of the MYC or HRAS pathways. We conclude that these gene expression phenotypes have the potential to characterize the complex genetic alterations that typify the neoplastic state, whether in vitro or in vivo , in a way that truly reflects the complexity of the regulatory pathways that are affected.

Published online: 18 May 2003, doi:10.1038/ng1167

Posted by erich at 02:57 PM
visualizing medicine

I've been thinking about this since taking the Radiology elective in February. Rays isn't easy, but once you've acquired the visual vocabulary, a film can tell you a lot very quickly, often instinctively. I've always been visually oriented and find it easier to assimilate large amounts of data if it is expressed in pictorial form. One of the treasured books in my library is Edward Tufte's The Visual Display of Quantitative Information. The best known figure from this book is the famous map of Napoleon's 1812 campaign in Russia by Charles Joseph Minard,


which succinctly summarizes the geographic path of Napoleon's advance and retreat, the Grand Armée's manpower, time, and temperature.

Work in the CODEx Project hammers home that we need to think of ways to present complex clinico-genomic information that is both comprehensible, and yields up important subtleties no different than the nuances a good radiologist sees in a mediastinal silhouette.

I would think this should be an additional focus for the Project as it matures.

Posted by erich at 10:50 AM
May 17, 2003
sars in taiwan

Taiwan reports its largest one-day increase in SARS yet. This is depressing--especially since my parents have rushed back. As head of a research hospital, my dad has to go back.

The sad thing is that I'm not particularly surprised Taiwan is having problems. Semiconductor foundries, Apple/Dell/IBM laptop manufacturing, and LCD screens are things Taiwanese are good at. Pre-emptively dealing with problems that don't have the seductive sheen of high tech or money is something the Taiwanese are not. There's also a culturally-ingrained tradition of ad hoc just-good-enough improvisation for solving problems (really no different than on the Mainland). With this sort of culture if it's just money at stake, you get breathless, percolating, euphoric economic drive; when it comes to something as mundane as public health, you get discouraging heaps of crap swept under the rug. This passivity is hurting them now.

On one trip over, I checked out an ambulance. It blew me away: basically a minivan with room for a gurney, no lifesaving equipment, and just a couple hooks for an IV. I asked my dad about EMS services in Taipei. His response: they don't exist. This is a wealthy country. There is no excuse. Here at Duke alone we have two Lifeflight helicopters and half a dozen trucks stuffed with equipment.

Compounding Taiwan's SARS crisis is treachery. For all I care, the head of Ho-Ping Hospital (the one who concealed early cases and essentially killed several of his staff) can burn in Hell.

The Taiwanese has successfully thrown off dictatorship and political repression, creating the only democracy in the Chinese-speaking world, they now have to throw off the repression of narrow-mindedness and passivity.

Posted by erich at 10:34 AM
modern women...

Front page of A bride in India, in the middle of her wedding feast, put her foot down and had the groom arrested when his family tried to extort an additional $25000 for her dowry just as the Hindu priest was warming up for the ceremony.

It's not as if the in-laws had already lined up not one, but two Sony TVs and home entertainment centers ("she specified a Sony...not a Philips"), two refrigerators, two air conditioners and a car.

Supposedly in the commotion that ensued after the in-laws' demand, the bride whipped out her cell phone and called the police (dowries, though practiced in everything but name, are technically illegal).

Strong work. With her newly acquired fame as "Miss Anti-Dowry", she's had 25 new engagement offers.

Posted by erich at 09:13 AM
May 16, 2003

Wellspring had these handsome soft shell crabs today. I took two home and plumped them in some skim milk. In the meantime I clarified some butter. My mom had left some cranberry vinaigrette--I probably would otherwise have made some tomato/basil, but wanted to use up what was in the Fridge--so I decided to use that with the crabs. After I dredged them with flour and ground pepper, and a sprinkle of kosher salt, I fried them in virgin olive oil and a couple tablespoons of the clarified butter.


Healthy? No.

Posted by erich at 08:08 PM

I just got my US Medical Licensing Exam Step 2 score...a Pass. I confess to a few disquieting dreams over the last month involving other interns co-signing my orders. Being in lab for several years + 9 days of desultory studying did not add up to much certainty when I walked out of the testing facility.

The report also shows the range of your performance in different subject areas. No surprise, the bands exhibiting less than scintillating performance were Psychiatry and OB/GYN. I suppose as someone who touched on Freud in his senior thesis, I should have done better in Psych, but the modern practice of psych is more about properly sticking the patient in the right diagnostic category, e.g. patient A must have symptom X for six months, plus 2 out of 3 symptoms in category Y, unless she has symptom Z, which allows us to reduce the six months for symptom X to 3.5 months depending on whether the phase of the Moon agrees with the traverse of Venus.

The nuances that differentiate "Schizoid" from "Antisocial" personalities just don't hold my interest.

As for OB/GYN, my performance is mildly surprising because I actually did the best on those questions on practice tests.

On the other hand, I have no idea what a gynecoid pelvis is.

In the end I'm glad that there are others who know and love this field.

What discipline did I do best in? Surgery. I also seem to have the strongest affinity for Immunologic Disorders as well as Diseases of Blood & Blood Forming Organs.

Interesting that going into the test I held that Step 2 represents the last time that OB/GYN and Psych can get their claws in the tens of thousands of medical students like me who will never again have to deal with OB/GYN and Psych other than calling a consult. It was exactly those two disciplines that got me--even left a scratch or two.

But I'm free now.

I should cackle diabolically, but it's hard to do in text.

Posted by erich at 04:00 PM
May 14, 2003
firing blanks?

David Edelstein's review of Matrix Reloaded seems to confirm my fears about the Wachowskis' ability to follow-up. Of course it's a review and its subjective, but it's nicely written and very witty (had to chuckle at the references to Jayson Blair and Cornel West). Not that I've been waiting, bereft, for the return of Neo, but I harbor the simple hope that it will be good entertainment. From Edelstein's review, it sounds like a technical masterwork without the ontologic mysteries of the original.

I was/am hoping that the Wachowski's would come up with something clever and mind-blowing, which Edelstein suggests they haven't.

I harbor the subversive opinion that Groundhog Day actually beat the original Matrix to the punch. How does the acne-scarred, sarcastic Bill Murray measure up to the smooth-planed Keanu? Punxatawney was no less an inhabited video game than the Matrix. By the second third of Groundhog Day, Bill Murray is a super hero. Since he knows exactly how the game works, he can subvert fate by being in the right place in the right time (and even impress Andi MacDowell by quoting Baudelaire). For a dingy little town in Pennsylvania, Bill Murray might as well be "The One".

The dramatic tension in The Matrix came from Neo not being aware of his potential and our slowly realizing that those hours playing CounterStrike on the internet weren't very far from Neo's reality. After that, we could start wrapping our minds around the geek fantasy that facility with CounterStrike or Quake might equate with our abilities in some social universe. So how do you restore that sense of revelation when Neo can do everything already? Bill Murray got so bored with the process that he spent part of the movie thinking of creative ways to off himself.

I guess I'll have to see for myself.

Posted by erich at 08:34 PM
chopin piano concerto no.1

I'm listening to it right now. Since accompanying this with the HRO many years back, I probably haven't heard it more than a couple of times. I don't think I differ from many in preferring Chopin's solo piano compositions. His orchestration isn't innovative in any way--it feels like a concerto plain and simple. In spite of its lacking compositional innovation, there's still that feel of a concerto.

It's hard to convey what I mean. For me there's a sense of drama with a concerto that's probably very much tied up in my identifying with the soloist: the subdiaphragmatic whir of butterfly wings as you stand in the wings after the orchestra has finished tuning; how the heavy velvet drapes bow outwards as you walk briskly onto the stage, the hardness of your soles against the wooden stage, eyes and the expectations of the audience and orchestra focused on you; the conductor turns to you, baton poised, eyebrow raised--a silent "ready?" A stray cough echoes through the concert hall; an indrawn breath, and the orchestra starts...

Posted by erich at 06:05 PM
more lancet paper news linkage

This from US News & World Report. This writer really gets what we're striking for with the paper. He prefaces his piece with the dilemma clinicians face when they have an effective, but toxic drug that works for a few patients, but they don't know which patients. So they have to just go at it trial-and-error, exposing some patients to toxicity without benefit in the process.

What we hope our method will help people do in the future is to help patients and physicians understand the "which?" question.

Posted by erich at 01:02 PM
learning the ropes

One interesting component of the work I'm doing is that it exposes me to the process of thinking about how academia and the commercial arena interact. If anything we're doing is going to impact people, it has to be "out there". And to be "out there" means engaging the hurly-burly of commerce.

During the boom, still being in school, I felt sidelined (and even thought about blowing it all off and exploring working in VC). Sitting on the margin was a blessing in disguise. All of the era's hubris, overstating/overreaching, are lessons learned from afar.

So now we're working slowly and methodically to build what we have into something that's good for patients, that we can be proud of, and that we can put before the public in a manner that treats all partners/collaborators equitably. I'm sure we'll make mistakes and stumble along the way, but as a lifelong student, I'm willing to treat all experience as learning experience.

Academics tend to look askance at business, but the gears and levers for getting things done and truly impacting society are--for better or for worse--commercial.

Posted by erich at 10:18 AM
May 12, 2003
more anticipation

Posted by erich at 01:14 PM
May 11, 2003

Snagged my MD diploma this afternoon. In the last week I've gotten two Duke blue binders to put in the safe. I'm not the kind to hang my diplomas on the wall, but I wish they'd make them bigger. My undergraduate one is an elegant and big affair. These specimens are 8.5" by 11". Can't they splurge a bit on the lambskin? I could run-off something like this on my printer! While I robed up for the Hippocratic Oath ceremony, I couldn't abide the thought of donning that polyester extravaganza today. Probably didn't get my money's worth for the rental. Oh well.

Posted by erich at 02:21 PM
May 10, 2003
oath taking

Last night was the Hippocratic Oath ceremony. A moist May evening. As we walked from the South Clinics out to Davison Quad, the Chapel bells were ringing most academically. My polyester PhD robe was a puffy, hot, and Duke blue affair. I tried to wear it as little as possible with the heat until our pre-ceremony Class Photo session forced us to line up on the Old Chem steps for our digital firing squad. Several years ago, the class that I felt the most emotionally connected to, my entering class, graduated. As a late appendage, I couldn't feel the commonality of shared experience. The class photo I have in my room is that one.

The actual ceremony is a formal affair where the graduating seniors line up alphabetically to file into the Chapel, its organ precessing, in something akin to a marriage ceremony. The Chapel was quite full and there was David, Pearl, and Nathan on the right, Tyler and Jake hollering for my attention, Mme. Rosse waving from up front on the right. On these occasions Duke was so smart to have built a "Chapel", though as Dr. Halperin pointed out in his excellent address, it really is a Cathedral aside from its lacking a bishop.

They had us seated in rows in the front, and Dr. Halperin thundered away on the millenia-old tradition of medicine, the etymological origins of our hallowed tradition, even a close reading of the actual Hebrew of Exodus 3, and that what it means to stand on sacred ground. His interpretation was that the intent of the passage was that the ground under your feet is sacred wherever you go; thus, standing before a patient in the clinic, we must understand our sacred role. The address was a good one.

We recited the Hippocratic Oath (of course one of its modern versions) and each went up to the altar to be hooded and receive a scroll inscribed with his name and the Oath. I felt sorry for the audience having to sweat through 87 of these. And now I'm wedded to my calling.

After the recessional (more traditional wedding music, Vierne), we emerged into the soft night. I found my father and spoke with Nancy Major, easily the most warm-hearted attending in the hospital, Jake and Tyler gave me a picture frame Jake made with their photos, insisting that I come babysit them soon. Tyler comically donned my cap.

Joe and Mike were there and much appreciated for their attendance, and the Rosses where there to witness the remarkable occasion that their neighbor had actually done with school.

Posted by erich at 04:57 PM
lancet paper news outlet linkage

Here are some newslinks related to the Lancet Paper:

From the BBC
From Wired News
From Reuters via Yahoo
And Reuters Health

Posted by erich at 08:48 AM
May 09, 2003
breast cancer genomics paper

Our group's paper is released for publication today. The associated commentary by Sridhar Ramaswamy and Chuck Perou is cogent.

My initial response to their commentary
They query the use of metagenes (they call them "highly abstracted structures") to summarize the impact of multiple genes, suggesting that treating genes individually is enough, and that deconvoluting the roles of individual genes from metagene data is difficult--I'd venture that (1) metagenes reduce noise; (2) more importantly, they reduce this noise in the context of discrete biological functions, i.e. all analytical techniques aggregate genes in some manner, what we do is distill and heighten the signal/noise ratio for genes that share functional associations; (3) metagenes actually simplify the process of understanding how individual genes interact within functional roles because they allow us to prioritize and estimate the impact of individual genes.

They also ask what the point of predicting lymph node status is if it's an "imperfect surrogate", but the point is that lymph node status is currently the single best clinical prognostic indicator, so (1) shouldn't we try understand genomic data in this context? (2) isn't there pure scientific interest in understanding the metagenes and biology involved in lymph node status? and (3) what about those patients who within a narrow window of time at workup are lymph node negative, but are about to convert to lymph node positive? Isn't it of use to be able to identify these patients?

Finally, they off handedly say that out-of-sample cross-validation "generally overestimates" accuracy. I'm not so sure this is a proper generalization to make. If we "locked" our predictive model and just cross-validated the samples, perhaps. But we cross-validate not only the samples, but the model. This is about as stringent as one can get. Naturally we are actively augmenting our sample size and hope to be working with thousands of samples in the near future.

Reasonable questions. Perou and Ramaswamy do find common ground in that gene expression data all point to the fact that metastatic potential is present in primary tumors.

Posted by erich at 08:33 AM
constituent service

The NYTimes has this article on Elizabeth Dole's first few months in office. Apparently she has spent all of her time engaged in serving her state rather than parlaying her connections and position to maintain a high profile in the Senate. It seems that many opponents are befuddled and even disappointed by her going to ground. The accusations of Dole being a carpetbagger hardly apply when she has turned down interviews in every major print and TV outlet and spent all of her freshman time visiting tiny obscure hamlets throughout the state. She only consented to her NYTimes interview because it would take place in Camp Lejeune.

After her bravura performance at her husband's nomination for the Republican ticket, I confess that I thought her a bit slick, and calculating; but her doing the seemingly mundane gruntwork of taking care of her constituency seems to force a reevaluation. A political consultant was left saying, "She hasn't done anything that Democrats could much attack here, much as we'd like to." Another disappointedly wondered if Dole was "ever going to move out her husband's shadow?" People forget that Bob Dole was respected on both sides of the aisle because he was an excellent Majority Leader who got the job done rather than cultivating political division, helmet hair and inserting both feet in his mouth. Apparently, his advice to his wife has been to do the same: do the job you are paid to do--that means getting Federal recognition of the Lumbee Indian tribe, working on a buyout of tobacco farmers... None of this the heady stuff of national politics for someone whom many considered the front runner for the Republican presidential nomination until W. came to prominence.

Several years ago, Henry Louis Gates wrote a glowing, even hagiographic profile of Dole in the New Yorker. to the point that it looked like she had his vote if she ever made it to a national election. There's probably more than meets the eye.

Posted by erich at 08:04 AM
May 08, 2003

Yes. Just a few weeks.

Posted by erich at 11:53 PM
making things personal

Dean Williams, Ralph Snyderman, & Huntington Willard published this editorial in Science. The leadership in the Medical Center has chosen the concept of "individualized" medicine as a rallying point for research. From the angle of genomics, I published this editorial in the Chronicle, December 2001. I think such an approach will be the important next step in medical care. Everyone is very aware that assigning a patient to one prognostic bin versus another can be arbitrary. A woman with breast cancer and 0 positive lymph nodes is treated differently than a woman with 1 lymph node. These sorts of tools are very "low resolution"--on average they work, but there is significant room for improvement. And because these bins are large, a patient isn't treated as a distinct entity, but as an undifferentiated member of a bin. What we're harnessing genomic information for is to get away from discontinuous categories and move into a continuous space that allows us to predict a woman's individual prognosis as a point estimate for her, not as a member of a group.

Ultimately, this sort of individualization will revolutionize how we prescribe drugs, how we determine cancer treatment protocols. It will even allow us to "recover" drugs that had been shelved because on average they failed to work for large population, but we will now be able to tell specifically for whom that drug is actually efficacious.

Patients instinctively recoil from being treated as merely being a part of a population, they want to be addressed as a singular person with singular health characteristics...thus the wackiness of plying Whole Foods' herbal remedy aisles, which is symptomatic of Big Medicine's failure to answer personal needs.

Next week, our paper which demonstrates some first steps in this direction for High- and Low-Risk breast cancer patients as well as for predicting lymph node status will be appearing in The Lancet.

Posted by erich at 02:31 PM
May 07, 2003
Andrew Firestone

...Is an idiot.

Posted by erich at 10:02 PM
this is progress

I've ridden in one of these. Our college orchestra's Eastern European tour included a sojourn in the soon-to-be Czech Republic. I remember my host family arriving in this clattering, plastic car, considerably noisier than my neighbor's John Deere rider mower. Now, according to The Guardian, a company wants to perpertrate this conveyance on Africa, formerly known as the Trabant, and dub it the "AfriCar". Its chalky, slightly fibrous and sun-bleached red finish left an impression, but I didn't realize that its body is made from "a blend of phenol [carbolic acid] and compressed cotton named Duroplast", or that its rubber insulation is "prone to melting and 'leaking like a sieve'". The advantage is that it's "unbreakable", and, assuming the impossible happens, "the car was designed for 'any roadside mechanic to fix'".

Who needs a Hummer?

Posted by erich at 01:17 PM

I had read the novel a couple of times in Middle and Upper School. Filtered through those adolescent impressions, in spite of its undoubted importance as a critique of totalitarianism, 1984 had always felt like science fiction to me. It's suffused with cold, and inhabited by the Byzantine apparati of oppression. This always requires erecting contraptions for the sake of atmosphere and plot that just feel elaborate and busy to me. I wonder if going back to it as an adult might affect my perceptions of the book, whether the vision of a prole singing would impact me more than its steel and concrete drabness.

Interestingly, in his introduction Pynchon calls the internet "a development that promises social control on a scale those quaint old 20th-century tyrants with their goofy moustaches could only dream about." I find this odd and paranoid, not because of the accepted wisdom that the internet is a democratizing influence--that is a cliche perhaps--but because Pynchon's predilection for rooting out secrets and subversiveness should have him reveling in a hurtling etherworld where hackers deface government and corporate websites, war-chalkers root out wireless networks, bloggers aggregate and debate. The WASTE symbols that Sixties college students used to scrawl in university bathroom stalls reminds me of the chalk iconography of war-chalkers. Thomas should get out on his computer a bit more.

It's ironic that the late 20th Century dystopian vision of a dehumanized police state with its accoutrements of technology is entirely wrong about the challenges that confront us now in the 21st. Rather than Progress running roughshod over basic human rights, it's reversion to Medievalism that threatens individual freedom. I wonder whether we are philosophically equipped to handle this challenge when for the last 80 years our energies were bent on critiquing and opposing Fascism/Stalinism/Maoism. This ideology of stentorian music, posters and statuary of musclebound mechanics and industrial production wrapped up in the trappings of Science and Progress is taking its last paradoxial and asphyxiated breaths in North Korea. If anything, science and progress have helped rather than hurt us. It's the wild-eyed fundamentalist plotting by firelight in a shack or a cave (whether he's in Afghanistan or Montana) with antiquated visions of what will buy him a place in heaven that frightens me more than the spider-like probes of Minority Report. I can opt out of registering my email address; I can turn off my cell phone. I can't prevent a 767 from colliding with my workplace. Not that we should be lazy about the issues of privacy that the wireless and internet world have spawned. But I think we have a vocal and vociferous enough group of technophiles defending the privacy border.

There are those that would argue that we shouldn't stop short with Medievalism being the greatest threat to modern society because the response to that threat entails aspects of that technological police state. And I'd be forced to agree up to a point. The biggest difference is that we have mechanisms for self-examination and public debate about whether we overstep in our desire to protect ourselves. For those captive in the Middle Ages, there is absolutely no legitimate and protected manner to make this critique.

Though Orwell pitched his story 40 years into the future, what he was really responding to was the here and now; i.e. what he'd seen as an disillusioned Republican in the Spanish Civil War, the brutalism of Hitler and Stalin. There is nothing really prescient about 1984. And, as Pynchon points out, he missed the boat on fundamentalism. be continued...

Posted by erich at 10:23 AM
May 06, 2003
lot crying

Whoa. Thomas Pynchon has written an introduction for a new edition of George Orwell's 1984. And it's excerpted in The Guardian. Thanks to for heralding this (via

Posted by erich at 11:46 PM

The Washington Post has an excellent report on how the CDC is responding to SARS here. The Post seems to be publishing more articles that are straight reportage without much of a polemical bent. There were some very well-written and atmospheric reports from their embedded reporters during the Iraq conflict, and this article seems to be more of the same.

I really enjoy pieces that give you a feeling for the texture of life in a milieu you haven't been exposed to, whether it's a firefight on a road in Iraq or the aroma of new "furniture carpets, and their heady hydrocarbons" in the new Emergency Operations Center at the CDC.

As someone who's seen both sides of the Pacific, one thing I see Americans do well in comparison is respond to crises. There's a deep vein of pragmatism in this country that results in competent specialists like Virginia's Fairfax County Urban Search & Rescue Team that's been designated by USAID as a response team for disasters across the world. Indeed, when Taiwan had its earthquakes a couple of years ago, the VATF were the ones to show up and help them deal with search and rescue. Same with the CDC. I'm sure they're not perfect, but if I had to choose between China's and ours, the choice is easy. Walk into a Home Depot and you'll instantly find dozens of people who know how to do a job right.

Chinese culture tends to emphasize getting something to just work by any means, whether that entails doing it the right way, or doing it in a way held together by bubble gum and rubber bands. There's also a tendency to sweep things under the rug that cannot be handled ad hoc, as has been the case with SARS.

China clamors to be accepted on an equal basis in the international community, but continually undermines itself with amateurishness. Their holding the plane and crew after their incompetent fighter pilot collided with a US military observation aircraft, their hesistancy to engage North Korea in discussion (until recently) with the US, and their failure to report and respond to the initial stages of the SARS outbreak are not actions of a modern nation aware of its obligations to the international community. If they want to be a WTO state, they ought to act like one. Ironically, their actions have probably cost them more economically than if they'd openly acknowledged this new disease from the beginning. They're a superpower only by virtue of their being a nuclear power and their massive population.

At least they're letting the WHO go to Taipei (which has been a major politcal bone for a long time).

Give me the Poles or Czechs any day.

Posted by erich at 07:48 PM

Columbian Mesa de los Santos, ed to setting 6.5, well into the second crack. A wisp darker than a Full City . I'll drink some after dinner.

Posted by erich at 05:12 PM

Whenever I ran into an attending or anyone medical-related, they'd ask me "so what are you doing before residency starts?" Lacking a strong response to that question, I began thinking about what I might do. Should I brush the dust off Tyler2, my trusty sea kayak? (American history geeks will understand the reference). Perhaps take it, filled with a week's worth of food, to the Outer Banks and explore salt marshes and haunt Portsmouth Island? Maybe if I wanted a more low key break, rent a house in Corolla, or perhaps Bald Head Island, and just read and write. Then there is the Offshore Sailing School class in Tortola, BVI. Bora-Bora even flitted briefly across my mind.

In the end I decided to do what 80% of honeymooners decide to do: go to Hawaii. Since I'm newly-wed to the life of a surgical intern, why not? So I've appropriately drained my parents' Frequent Flyer fund (United Airlines' Premier Executive Elite Gold Member Special Private-Line-Only reservationist was rather surly in spite the rigmarole of a special treatment and a special 1-800 number), pre-paid the hotel and car rentals, and am on my way to Kauai and Maui. I plan to be surrounded by nubile hula dancers necklacing me with fragrant orchid leis and plying me with umbrella drinks and Kona coffee (I'll probably arrange a day trip to the Big Island).

Don't let me down Hawaii.

Posted by erich at 08:28 AM
May 05, 2003

I tripped across William Gibson's blog and found an interesting comment here. I haven't wrapped myself in the blog universe, but agree with Neuromancer's author that the relaxed, colloquial nature of blog writing doesn't meet the stringency of what we traditionally call good writing. People tend to be conversational and self-conciously irreverent. But then when I want eloquence I don't go fishing for it on the internet. I wonder if Shakespeare would have kept a weblog.

For instance, I see the F-bomb dropped a lot. I've nothing against the F-bomb, and often find some amusement in its employment. I think I'm still subliminally shocked by its appearance in print and get that thrill of taboo blaring out from my laptop screen, but it's a cheap thrill.


Posted by erich at 02:11 PM

Since Saturday I've been spasmodically fiddling around with utrinque paratus. Evidently, this site has been neglected for a long time, but here I was with time on my hands, the realization that I was 0.63 versions behind with Movable Type, an obsessive desire to figure out a way to post on utrinque paratus what I'm listening to on iTunes currently. Amazing how a day can skulk silently away while you're hacking on the computer. I can see his bald head rounding the corner.

It's fun. Bizarre how I can take a photo, manipulate it, create a new banner and have it posted on a server based...where?--I don't even know where my host is based--and have it lighting pixels on the internet in 10 minutes.

What a different world internet publishing has wrought.

Posted by erich at 01:20 PM
May 04, 2003
kung-tunes primer

It really is straightforward. First download the application from here. Then check out the author's instructional blog entry.

This is oriented to Movable Type users who have the TrackBack function built-in. Normally this is a cool way for weblogs to easily refer to one another by pinging to a .cgi script. The latest version of Kung-Tunes takes advantage of this by using the "HTTP POST" method to ping iTunes info to the mt-tb.cgi in your Movable Type installation.

So, once you've downloaded the app, go to your Movable Type website's control panel and select the appropriate blog (assuming you have a couple). In the sidebar on the left, select "Manage/Categories" and create a new Category called "nowplaying". Click on the "Save" button and wait for the screen to refresh. Now "nowplaying" is added to the list of Categories you already have running. Click on the "Edit category attributes" link next to the "nowplaying" field. Here you can add a description of "nowplaying", but the key is to scroll down to the section "Incoming Pings" and click the "On" radio button for "Accept incoming TrackBack pings?" Click the "Save" button. The screen will refresh and you'll see below the "Incoming Pings" section a new section headed "Trackback URL for this category". In the box below will be an URL that looks something like:
This will vary depending on how you installed Movable Type, and the numeral after mt-tb.cgi will vary (I think). Copy this URL into your clipboard.

Now go back to your blog's management sidebar and click "Templates" and select "Main Index". At this point, for prudence's sake I'd copy the original "Main Index" template to TextEdit in case you goof. Now Kung-Tunes's author doesn't recommend just cutting and pasting the HTML that inserts your iTunes track info directly into your template, but assuming your template HTML is straightforward (e.g. the standard column layouts), this is quite doable. I have a heavily modified version of one of the templates that come standard with Movable Type and knew exactly where in the right hand column I wanted to insert my "Now Playing" section. Here's how it looks for me:

<div class="sidetitle"><br /> Now Playing<br /> </div></p> <div class="side"> Below are the tracks I'm currently listening to on iTunes 4:<br /> <p><MTPings category="nowplaying" lastn="5"><br /> <$MTPingExcerpt$><br /><br /> </MTPings><br /> </div>
So the "MTPings category" is the new "nowplaying" category we created where "lastn" refers to the number of most recent tracks you've listened to.

Now open Kung-Tunes and select the "Preferences" panel from the Kung-Tunes Menu. Make sure the "Upload" tab is selected and that the "Upload Method" is HTTP. Paste in the TrackBack URL into the "URL:" field. Make sure you delete the "http://" from that URL. Kung-Tunes automatically inserts this and otherwise you'll be left scratching your head why you keep on getting "Type 6 Errors" whenever you try to upload iTunes data to your blog. It took me an hour to find this boneheaded error. Click the "Save" button. As is advisable with any program, if a "Console" is available, open it if you're having trouble; it's much easier to spot silly syntax errors if you can see what the application is telling the server.

Otherwise, the defaults in this panel are fine. Next, select the "Formats" panel from the menu and delete everything from each of the fields and paste

title=itunes&url=foo&blog_name=myblogname&excerpt=^t by <b>^p</b> (^a)
into the "Format for currently playing track" field. You can play around with the tags later if you want to customize how the tracks are formatted. Also make sure to enter your own blog's name where you see "blog_name" above. It's the name you specified in the Movable Type control panel. Click the "Save" button.

You should be set now. Fire up iTunes. Kung-Tunes should show your track info. Click the "Upload" button and a small bit of text with your track info that's saved in your /Library/Preferences directory should be sent to your Movable Type installation's "mt-tb.cgi" and published on your website.

This primer isn't radically different than the author's own directions, it just clarifies some points.

Posted by erich at 03:27 PM
now playing

As many have, I finally installed Kung-Tunes, a slick Applescript application that gets track info from iTunes and pings your blog so you can automatically post what's playing on your computer on your website just as I am now running in the lower right corner of this page. It cleverly uses the Talkback function in Movable Type to do this without php or javascript (which are both doable using Kung-Tunes as well).

Since it took me a bit of time to get it to work (more out of my own dunderheadedness) I thought I'd add a brief tutorial here.

Posted by erich at 02:53 PM

A week from today I'm done. It only began to sink in when I finally tracked down my AWOL PhD diploma, and picked up my robes, hat and hood. Up to this point as I've finally kicked past the last few milestones, I've tended to view them with some sense of anticlimax. Last August I didn't make any real effort to bring my family to my PhD defense and it was Joe who encouraged them to come. For next week's MD activities, I really would have been happy to sneak through the whole process, but alas, they're here.

Walking across the Quad, my diploma in one hand, a hanger-full of academic regalia in the other was the first intimation that I've actually done something. These are all symbols, but it's finally palpable that I've finally laid down the foundation, and now it's time to get to work and build upon it.

March 20th was a rain-spattered Match Day. I'd scrubbed out of a breast reconstruction case, put on my white coat and tie and wandered down to the Searle Center as thousands of Duke seniors have done on the third Thursday of March. There were only a few people yet. Most were in street clothes having had the day off. I suppose I should have asked for the same, but sitting at home, bored didn't appeal to me. Having been only a late appendage to this class made the day less momentous I think. I only knew the few people I shared rotations with and a few holdovers from the class I had TA'd in Medical Genetics. I ate a lot of chicken tenders.

Within ten minutes the room was milling with nervous 4th years. It does seem that everything about medical education encourages high sphincter tone. We're either apprehensive about our grades, which medical school accepts us, whether we survive medical school, then the Match, then residency, the fellowship, the Boards, on and on. There was Suzette in a Burberry hat, there was Patty saying that she felt like projectile vomiting, Daniel waiting for Emily to show up, Sheleika looking far more relaxed than she had in the SICU, Shy to witness the proceedings and provide moral support. They had these really huge chocolate covered strawberries. I had about five--they were in front of me; I couldn't help it.

And then here I was, putting a dollar in the pot, shaking random hands and collecting the envelope that would tell me where I would spend the next at least five years. After the 87th person had been called was the signal to open our envelopes: "Contratulations, you have matched!" in bold on a laser-printed page from the NRMP, "General Surgery, Duke Univ Med Ctr--NC". What I'd wanted.

It all begins soon. I confess that I'd rather live and work in San Francisco or Boston, but work in the CODEx Project is so promising, and I'm to have some sort of faculty position in the Computational & Applied Genomics Program of the Institute for Genome Sciences and Policy, that leaving seemed counter-productive. Anyhow, I actually don't mind the infamous White Pants of the Duke surgical Intern and Junior Resident. We may indeed be the last to require them, but aside from having to find white boxer shorts, I like the idea of minimizing early morning decision making.

Posted by erich at 10:16 AM