Wednesday, November 24, 2010

In which I come up for air, and play with github code snippets

I've been buried recently in job hunting, after having moved from Manhattan to San Francisco. Long story, and if I don't know you, I'm not likely (nor willing) to bore you with the details. Anyway, this effort has generally required I demonstrate my programming chops - over, and over, and over - and in the process of doing so, I've found I need to document bits and pieces of environment configuration, command line one-offs, and other miscellany. I'd been looking at Blogger plugins, or plugin hacks involving 3rd party javascript packages such as SyntaxHighlighter, but have found a more elegant solution: publish gists on github and embed a one-line javascript directive, e.g.:


If you can see this, I am happy. And by "see," I provisionally mean "see with your eyeball(s)," since I can't guarantee your RSS reader can handle this. But it's a start.

Friday, October 29, 2010

Memebusters Episode 1: "Einstein Never Said That"

A few days ago, I posted a short, awkward, mainly Epic Fail version of the "Bible's broken" scene between the characters of Shepherd Book and River Tam from the Firefly series. I was playing with the text-to-video tools from Xtranormal, and no one who's not already a fan of the original Joss Whedon series should pay attention to my shaky first effort: it sucks. But it was an interesting learning experience. I had in mind to treat of some personally interesting pet peeves, and the first I chose was a particularly annoying meme, the "Einstein Buddhism endorsement" myth. Like many of my rants, this one has its roots in a rather insalubrious exchange with a person I think should have known better. Yes, I'm being catty, but I think my reaction was merited. There's too much deification of interesting historical figures by those who haven't the slightest understanding of the actual accomplishments of their rhetorical sock puppets, and Einstein is a common victim in these misappropriations. Here, then, for your delectation, "Memebusters Episode 1: 'Einstein Never Said That'":

Monday, October 25, 2010

In which I commence exploring the Uncanny Valley firsthand

I've been fascinated recently with the idea of text-to-video translation tools. Having seen some amusing YouTube product using the Xtranormal service, I figured I'd try my own hand at it. These tools are usually employed to render a couple of characters bantering lines fed into a speech and video rendering engine. Despite some rather obvious limitations, the results are often entertaining, so the output is generally not unpleasant to watch. I figured I might be able to bend these tools to my own ends, maybe to implement some kind of videolog presentation of lab protocols a là JoVE, the Journal of Visualized Experiments. I spent a bit of time trying out the free online version of the vendor's toolkit, an annoyance forced by the lack of a Mac OS X version of their client tool State.  I'm not sure how Uncanny a Valley I've dug with this first attempt, but here it is anyway: a quirky little adaptation of the "Bible's broken" scene between the characters of Shepherd Book and River Tam in Joss Whedon's television series Firefly.

Friday, October 8, 2010

On the importance of sequenced model organisms, & a crude taxonomy of their users

I've carved out some time recently to meet and interact with more of those people, like myself, who identify with the nascent DIYbio community. It's been a genuine pleasure. Some of this networking has happened at summer conferences, such as H+ Summit held at Harvard and Open Science Summit at Berkeley, as well as at smaller, less formal events, including one held earlier last week at a downtown Manhattan design studio used by an intriguing project in Columbia's architecture program. But as with most hackerish subcultures, my richest interactions have happened between face-to-face meetings, mostly on the diybio GoogleGroup mailing list.

I've had to struggle with a few things, and do a bit of Blind Man and Elephant feeling around, but I'm starting to develop a model of the beast in my head, one probably partially right and likely largely wrong, since the beast seems to be growing on a bauplan not comporting with a developmental model I, or anyone else, fully understands. So far, I've provisionally identified four rough categories of aspiring DIYbio biologist:
  1. The Educator
  2. The Hobbyist
  3. The Entrepreneur
  4. The Independent Investigator
The Educator's interests seem to lie in outreach, addressing the shortfalls and broken-ness of an embarrassing government public education system. He supplies passion, expertise, and - sometimes - equipment to cash-strapped science departments. His challenges are manifold, but his cohort's solutions are remarkable, such as a proposal to supply schools with PCR thermal cyclers at a low enough price point to make economic sense for a teaching module lasting a couple of weeks: a cheap, reliable, plain-Jane $200 cycler. Proposed devices may look like prototypes, but they're vastly preferable - and more easily justified - to the average run of knackered $800 commercial units sold by indifferent hardware clearance houses on eBay. And there are simply some approaches to teaching best seeded by outside ingenuity.

The Hobbyist is, of course, the consummate do-it-yourself'er: she's the DIY purist, and her motives are much more easily evident. She wants to tinker with everything at every step of the way, preferring all her own metal bending, DNA electrophoresis gel compounding (plain Jello gelatin is not to be laughed off, folks), and like one of those fabulous self-taught cooks who prides himself on where to source the best spices in New York City's Astoria or Chinatown, she would rather source her wildtype flies from a neighborhood compost heap and her wingless flies from a pet supply operation, than order them from a standard supply house like Carolina Biological. She is a dedicated life hacker and preferential locavore, and she advances the art and brings to it a delightful panache.

The Entrepreneur has plans, and could benefit in the execution of them by having access to low-cost facilities and equipment for prototyping new processes. He's probably just out of college, or a career changer, or both, and his highest barriers entail access to gear & expendables for proof-of-principle exploration. He wants to try some things out, and he needs low-cost access to a friendly workspace where the owner-operators aren't waiting to pounce on him and lay claim to a piece of a potentially lucrative business model. In fact, he'd really prefer a social setting in which he can collaborate with others, and maybe get noticed by potential angels and other investors. You might want to hire this guy, or at least stalk him from afar and throw money at him at a vulnerable "Eureka!" moment.

The Independent Investigator is a guy with some academic grounding in the field, maybe some additional industry experience in (or on the periphery of) biotech or pharma. He might be between schools on a gap year or two. He might want to work on something his lab's PI doesn't find interesting, or in a facility which doesn't allow "Saturday projects" as as matter of policy. Or, maybe he does work with an enlightened team in fabulous facilities, but fears the conflicts of doing his own thing under the aegis of an NIH grant not scoped for his pet project: who owns the outcome? So he chooses to set up a bench at home, or in a shared space where he might do his thing unencumbered. Because sometimes, even professional chefs prefer to cook at home.

Now, these are rough, subject-to-revision taxa, and exist mainly to help me organize my own thinking on the matter. Like most early attempts at classification, they're rife with observational deficiency and confirmation bias, and I expect my categories to be revised over time as the community and I mature together. I find these roughly useful, though, especially when used in combination, to model the motives and project the needs of individual actors, myself included. The interesting fact is, most of those involved don't fit in one category to the exclusion of others. Many of us harbor the Educator impulse, that desire to share domain expertise with those who might join us. Who doesn't have a bit of the Hobbyist in them too, that drive to build one's own toys? And in this community, the Entrepreneur and Independent Investigator are often aspects of the same agent. This last embodiment feels most kin to me when I meet it at events, and he and she are my intended audience here, though others may also find what I have to say of some use.

For the sake of brevity, I'll refer to my own cohort as "3:4" per the list above. Having first met 3/4s in person, and 1s & 2s later, I'd erroneously assumed the majority of DIYbiologists to have had some formal education and lab background like myself, since these were the people I was meeting at conferences. I think I was wrong, at least if mailing list traffic is any indication of underlying demographics. At least, I don't have good reason to attempt a census of lurkers, because actively interacting participants are likely to be much more interesting than lurkers anyway.

Given the relative scarcity of type 4s like myself, then, I've seen some discussion predominated by 1s and 2s in which the topic of some model organism comes up, but without the usual banter about strains and other identifying information one would expect when talking to a collaborator in another group about a mouse, say, or a fruit fly: "We need a couple of weaned litters of Black 6 delivered to our barrier facility," or "This protocol's DNA extraction requires at least 50 wingless flies to be homogenized first." These organisms are inputs to experiments, and one can assume -from how the critters were sourced - they've had their DNA sequenced or, at the very least, to have been genotyped. For those who don't understand the difference between "sequencing" and "genotyping," by the way, some study is in order. To those working with E. coli, M. musculus, and D. melanogaster: you have the benefit of having fully-characterized, sequenced organisms available to you now. Be glad you were born in these times.

To those of us in the 3:4 crowd: unless you've planning on personally conducting the first reference sequence of the strain of whatever organism you're working with, you're need to use a strain with documented sequence. More often than not, you'll need the particular strain that commercial and academic investigators are using, because assumedly, you're either working on a problem which has its roots in the literature (but which no one's yet tackled) or you want to extend someone else's work yourself. For example, you may think you want to take a crack at the potentially lucrative MPrize yourself, but even if you accomplish its stated objectives and produce a spectacularly long-lived mouse, you are in for a bit of trouble if you started with mice of unknown provenance: how are you going to expect others to replicate your findings and take you seriously? You may be a short-term pop journalistic curiosity at best, but you'll have to start all over again with a known strain.

Now, if you sit squarely in the Type 1 or Type 2 camps, then acquiring your flies and mice in the urban wilds is not only perfectly cool, but could also be fun, and might make for some interesting field trips, as well as great stories to tell. If you're interested in educating yourself or others about Mendelian inheritance, for instance, it could be argued that growing up city flies in large lots and selecting interesting mutants with obvious naked-eye-evident phenotypes, then crossbreeding them, is exactly the way to go about doing it. I would agree, and might even strongly advocate that approach. The last thing one wants to do is tell an aspiring child biologist that in order to work with flies, her only option is to have Daddy order a box of them from The Fly Store. That's a great big Curiosity Hammer swinging its way down hard on her motivation, a tailor-made way to kill the notion that she can do the stuff herself. Not good, not recommended.

But if you're wanting to "do science" and generate data for hypothesis testing, you're going to have to follow a few of the same ground rules of science that obtain regardless of your background or institutional affiliation. OK, so I've ranted a bit about the requirement to work with known model organisms, but I've said Jack about the logistics of actually going about it. The short answer to that one, my friends, is that I don't yet have a full set of answers. I do have some ideas, though: the biological sciences department of your local university is not a bad place to inquire, at least when it comes to finding E. coli and flies. To anyone reading this: have you tried asking someone? I mean, it's hard not to breed flies, given non-demanding room conditions, and they do breed like flies. I've worked in labs where excess flies are available, and often have to be killed. Try asking around. And E. coli? Bring your own LB medium and they might inoculate a culture for you. Good to go. Mice, though, are a very different issue, one I hope to treat of later.

Added bonus: you might just develop a really valuable relationship with a local lab, one whose postdocs might not have enough grad student helpers; if you can earn a track record as a reliable volunteer "outsource experimenter," you're golden. Consider it seriously. I'm surprised to hear nary a mention on the mailing lists of this type of thing, so it's worth mentioning.

Over the next few weeks, I hope to learn more about how the community sources its model organisms, and expect to report my findings when I can make concrete recommendations. Stay tuned.





Friday, September 24, 2010

In which, unable to sleep, I inaugurate a mostly politics-free DIY biology blog.

I used to blog, some years ago. I wrote some couple of thousand entries across three self-hosted MovableType managed blogs. One of those blogs took over over my "free time." Eventually, I dropped the lesser blogs into the bit bucket and let their domain names expire. I continued posting to the remaining blog; in the fullness of time, I let that remaining one go dormant. My reasons for dropping out were varied: "I'm too busy" to "MovableType template maintenance is a pain in the ass" to "I'm tired of hand-crafting DELETE FROM statements in SQL to ream out the CMS' backing store of penis enhancement comment spam," but really it boiled down to this: I simply got tired of blogging. Less interested in writing, discouraged by lack of feedback, and bored at the generally spotty quality of interaction my posts elicited from the few motivated readers who'd stumbled upon my blog. It had become a chore.

And of course, "life happened," as per cliché: I left a good job at a great company in order to pursue a long-harbored passion for regenerative medicine. This meant giving up a comfortable income, a nice California apartment, and shifting as far east as continentally possible without falling off the drier part of the land shelf. I moved into a one-room university-owned apartment in a pre-War building, hit the books hard, slept little, and emerged a few years later in May 2010 with a Bachelor of Science degree in Biochemistry from Columbia University, along with a year's experience from benchwork done at the Genetics Department of Albert Einstein College of Medicine.

In other words, I was ejected from the undergraduate pipeline an impecunious-but-proud, dazed-but-happy, moderately well-grounded novice in a new field, after having achieved some measure of "seniority" in another. Pretty crazy, huh? But I hadn't been living the life I'd wanted to live, and in the words of Robert Heinlein's character Daniel Boone Davis from The Door Into Summer, "In the course of a long life a wise man will be prepared to abandon his luggage many times." I don't count myself particularly wise; smart, yes, but not necessarily wise. This isn't false modesty speaking, but rather a genuine expression of a fear I think we should all share: that of getting intellectually soft about the hindquarters. Smugly, dumbly, comfortably soft. I fear the Previous Investment Trap, I dread throwing irreplaceable time after unrecoverable sunk costs. So, I shook things up and did something different.

Now that I've paid back around three years or so of a rather impressive sleep debt, I'm back at it. The stuff I plan to write here comprises all that miscellany inappropriate for email, 140-character Twitter pseudohaiku, and Short Attention Span Theater postings to my own limited social graph on Facebook. I simply despise the thought of writing acquaintance-tagged Notes on Mr. Zuckerberg's centralized panopticon cocktail party cork board, only to have them disappear eventually down the Orwellian Memory Hole. No thanks, I'd rather throw some thoughts and work notes out into the wild, where I maintain at least some small measure of control over persistence of memory and referential integrity.

This blog pertains to my new life as a biologist. Expect little in the way of political ranting, except insofar as politics gets in the way of doing my thing, e.g. the occasional venomous rant against cretins who insist on imposing their pre-Medieval views of ensoulment on the rest of us and mucking with our stem cell research. That kind of thing, you know, standard fare for the life sciences and those who try to practice them unmolested. Here I'd rather publish my domain-relevant longer opinions, my code snippets and technology explorations, and an occasional book review.

That, my friends, is my characteristically long-winded way of saying, "OK, I'm blogging again, for what it's worth." Hope you like it. Feed me, hire me, buy me beer.