Wednesday, April 30, 2014

"...every one of them is slowly going mad."

I found this blog entry about how much programming sucks though Lifehacker today. If you were to ask me what the top problem with programming is, I'd say something to the effect of "programs are written for computers rather than for people" or "code is like a hammer you have to read" and so on. (You didn't ask me about programming but I'm going to tell you anyway.) I'd suspect that most of what makes code "bad" comes down to the inherent differences between computers and human beings. Humans tend to freak out without context while computers don't need reasons to get things done. Humans see value in hard work, so they'll hammer away at the same problem for weeks or years; a computer will do the same but we don't see any value in its long-term labor unless there really isn't a faster way to do it.

So, I think I interpreted the blog post differently than Lifehacker did. Their idea was "The good news, if you're learning to code, is you don't have to worry (too much) if your code is bad." The impression I get is that you just can't write good code unless you're also doing good writing. That means you're simultaneously writing for an audience with perfect comprehension and no consciousness (computers, that is) and for human beings with plenty of consciousness but far from perfect comprehension. That doesn't sound fun at all. The solution most programmers use is to not even bother.

A few choice quotes from the post echo these sentiments:

Not a single living person knows how everything in your five-year-old MacBook actually works. Why do we tell you to turn it off and on again? Because we don't have the slightest clue what's wrong with it, and it's really easy to induce coma in computers and have their built-in team of automatic doctors try to figure it out for us. The only reason coders' computers work better than non-coders' computers is coders know computers are schizophrenic little children with auto-immune diseases and we don't beat them when they're bad.

The human brain isn't particularly good at basic logic and now there's a whole career in doing nothing but really, really complex logic. Vast chains of abstract conditions and requirements have to be picked through to discover things like missing commas. Doing this all day leaves you in a state of mild aphasia as you look at people's faces while they're speaking and you don't know they've finished because there's no semicolon. You immerse yourself in a world of total meaninglessness where all that matters is a little series of numbers went into a giant labyrinth of symbols and a different series of numbers or a picture of a kitten came out the other end.

It's a good post. It's very cathartic.

Monday, April 28, 2014

Today's paper is: 
Legrain, P. & Rain, J.-C. Twenty years of protein interaction studies for biological function deciphering. Journal of proteomics (2014). doi:10.1016/j.jprot.2014.03.038.

I keep changing the reference format. This really shouldn't bother anyone but me.

The take-home message: There's a lot of protein interaction data out there! Contrary to popular belief, most of it isn't just false positives. Rather, most of this data reveals actual biological complexity. Proteins may just interact with more binding partners than we originally thought.

A few awkward points:
  • Both of the authors are experts in the field but are also employed by Hybrigenetics, a company providing protein interaction screen services. They don't advertise the company's services specifically so I suppose it isn't really a conflict of interest. I still grow concerned about such things.
  • The title sounds odd. Couldn't it have been "Twenty years of biological function deciphering by protein interaction studies"? I think it's the word "deciphering", mostly.
  • The entire review sounds a bit strange, actually. I assume it's a result of English as a second language. It does make some conclusions hard to understand, i.e. "Almost 2000 different proteins were analyzed over six time points, covering four orders of magnitude in terms of protein abundance. In those papers, the aim of the purification process is no more the isolation of a protein complex but just a way to zoom in a specific part of the proteome", which sounds dismissive.
  • More specific references to recent interactome-dependent functional studies would have been nice. As they mention, there have been thousands, but it continues to be an active research venue. It looks like the authors just forgot to add some references in other sections: i.e., "..in different cellular contexts (for review on affinity-purification coupled to mass spectrometry, see)".
These guys made an early protein interactome of Helicobacter pylori so I can't complain too much.

Thursday, April 24, 2014

I never seem to remember the name of the film Mafioso. It's the one with the Italian guy who gets shipped to NYC by the mob. It comes up in conversation every so often but its title is just so generic that it's hard to recall.

Wednesday, April 23, 2014

Here's a neat little story about frozen frogs. The basic idea is this: wood frogs freeze solid over the winter, though that's not a problem since they can keep ice out of their cells (otherwise, all those ice crystals can really mess up those frog tissues). They do it by using glucose as a cryoprotectant and a specialized glycolipid as an antifreeze.

The paper in the Journal of Experimental Biology is here.

Tuesday, April 22, 2014

Beats!

Music for today - like many on Youtube, this guy does very high-quality song covers. The difference here is that he covers electronic-heavy songs that would be impossible for anyone to cover without a finely-trained ear, perfect timing, and just the right batch of equipment. I usually prefer covers to move in entirely new directions (the lady and I have been musing about what a female-fronted cover of December, 1963 would sound like*) but these are impressive on a more technical level.  Here are a couple.

A cover of Lovely Bloodflow by Baths. That whole album is wonderful.

A cover of Windowlicker by Aphex Twin. He doesn't quite capture as much of the original's classically weird atmosphere but that may be impossible.

A cover of Lost and Found by Amon Tobin. This is the live version, for reference.

Courtesy of this Metafilter post.

*A good cover. There are some on Youtube but this is the best one I could find, if that tells you anything.

Monday, April 21, 2014

Hello there! You may have noticed this blog transitioning to more of an electronic portfolio. This means that it will be less of a collection of musings and more of a collection of musings with a good bit of Context included. The rambling journal entries are still a primary focus. Now, however, you can learn about me in the "About" tab up there. I've included my personal mission and Curriculum Vitae. There's also a tab for Research: this page describes my current area of study. More material is on the way.

Thursday, April 17, 2014

On networks of the social variety

It's a strange time when I'm feeling better about Facebook and more uncertain about more specific social networks. In theory, the Book of Faces really is the internet's Walmart: it's a place to get all your social interaction needs as long as your needs aren't too specific. It's great for small talk but actually quite inefficient for any kind of extensive dialogue (to see what I mean, try picturing your next Facebook comment thread as a real-life conversation. It gets pretty difficult in those 10+ comment rigmaroles). The same goes for Twitter.* Even so, there's something to be said for small talk. It's better than no talk at all.

The constant level of background chatter is also preferable to noise with a perceived sense of urgency. In real-life terms, I'm talking about ringing phones, ambulance sirens, or other warning klaxons. These are notifications designed to get your intention; they signal immediacy and the need for rapid action. That phone isn't going to ring all day and that ambulance may need the road you're on. A bit of that urgency shows up in email and social networking messages. In each case, we see a tally indicating how many new requests for our attention there are today. Most of those cases aren't even as urgent as a ringing phone. We can ignore most of what happens on Facebook without any cost.

I've been trying out Researchgate over the past few weeks. It's one of the smaller, more specific social networks I alluded to above. It came strongly recommended by folks in my personal development class. Researchgate is essentially a social network specifically for scientists. Each user has a profile as with other sites, but the focus is less on their biographic details and more on their scientific accomplishments (read: publications). The site is well-designed and leverages a more natural approach to networking than sites like Linkedin do. Connection suggestions are frequent but appear to be weighted in favor of real-world contexts (i.e., colleagues in your department, people in your field, or authors you've cited) rather than a largely generic network structure.

The mildly irritating part about Researchgate is the question-and-answer portion. Much like Stackoverflow, it's a forum for technical questions sorted by method or scientific topic. Unlike Stackoverflow (or even Reddit, for that matter), the moderation is minimal. Many questions are hardly questions at all. I can excuse the lack of English proficiency in many cases but it renders many questions impossible to interpret. In other cases, the phrasing is nearly perfect but the question asker fails to provide anything but the most basic details. The worst part may be the sense of urgency: these questions give the impression that entire careers are on the line. A wrong answer, employed in earnest, may waste weeks or even months of valuable research time. I know it isn't my fault if people don't know how to ask helpful questions or how to take advice from internet strangers.** Social networks are just enough of a potential waste of time as it is.


*I'd suggest that Twitter is the Starbucks of social interaction. It's ubiquitous, frequently noisy, and redolent with fashionable credibility, yet even the people who use it often don't seem to devote much mindspace to it. It's usually pleasant yet generally forgettable.

**This is the way.


Additional note to self: don't forget that Biostars exists. It's Stackoverflow for bioinformatics.

Tuesday, April 15, 2014

Today I found out about the grid.arrange function for R (it's in the gridExtra package). It's one of those things which would have been useful to know about years ago. I just found it on a Stackoverflow answer today. The function just takes plots and squishes them together into a single canvas. It's easy, too: just enter
grid.arrange(p1, p2, nrow=2)
and you have everything in one place. The results look great with ggplot's faceting, too!
The extensive package library is absolutely one of the nicest things about R.

Monday, April 14, 2014

Despite the absolutely ideal weather this past weekend*, I spent a small chunk of time beginning to learn Ruby. The programming language itself seems fun and friendly; it's billed as a conceptual hybrid of Perl and Python with the object-oriented nature of something like Java. There are a few reasons why I'd like to learn the language:
  1. Learning new programming languages enhances understanding of the others. My Perl and Python can always be improved.
  2. I've seen a few bioinformatics-related job postings lately which specifically mention the desirability of Ruby experience. They're usually talking about Ruby on Rails and its use in web development. 
  3. It's either that or go back to learning Haskell. As far as learning a new language goes, most programmers seem to take Ruby seriously, at least.
I do have a few concerns.
  1. Is Ruby (or Rails) becoming obsolete? There's clearly room for a great diversity of frameworks in the scope of web application development. I'm also not qualified to answer such a question. Programming is just one of those areas where a newly-learned language or skill can quickly become a dusty relic. Every framework might as well be Esperanto in the long run: they all have their strengths, even if the only strength is facilitating communication between the right people.
  2. Will I ever really need to know Ruby (or Rails)? This question is even more difficult to answer, but I'll give it a try: "probably not." There will always be someone with better code than mine and it may even be in a different language or even an entirely different framework (for all I know, even scientific applications will be most frequently run in a mobile OS in a decade from now). 
Learning for fun seems like the best plan of attack. I can do it at my own pace without worrying about how applicable it is to Doing Bioinformatics. I'll just have the background to know it when I see it. 


*Here's a shot of the grounds at the University of Virginia.
Brooks Hall.

Friday, April 11, 2014

I've been enjoying the short "Futures" science fiction stories in Nature since they started. They tend to be concise, punchy, taut, and perfectly suited for the science-savvy audience. The most recent entry, "How Kameron Layas rode out the crash", is an ideal example. It breezes through some post-cyberpunk hallmarks without getting bogged down in anything but the most critical details, yet hints at a fully-fleshed periphery. It might even make a nice novella.

The author is Rahul Kanakia. The guy studied economics at Stanford but now writes SF. It's a natural progression but not a common one. He's also just a tad older than I am, a sure fire trigger for the "what are you doing with your life when this guy has already written a bunch of novels" effect.  This is alleviated somewhat by his great blog.  It's very analytical and unpretentious. If I ever decide to give up science for science fiction, I'll have to follow some of his examples. He's not big on advice, though.

Wednesday, April 09, 2014

This is the Donnelly Centre for Cellular and Biomolecular Research at the University of Toronto.

That's a nice-looking building. It was built in 2006 and really captures a particularly minimalist glass-and-color-burst aesthetic (the building reminds me of a larger, more technical Pompidou Center at least, and that building has been around since 1977). I'm curious about how energy-efficient all that glass is.

I think I just like labs with plenty of window space. It may be that the natural light cycles provide a constant reminder of the power of omnipresent natural processes. Or, it may be symbolic of the open exchange and transparent review characteristic of an ideal scientific process. The colors are nice, too. They're always a great way to find your way around a building and orient yourself.* Little details like that can really make the process of science much more enjoyable and efficient.


*In the absence of colorblindness, of course. Perhaps each area of a building could have a corresponding color and pattern.

Tuesday, April 08, 2014

I felt very caffeinated today. For me, the line between "fairly awake" and "very caffeine-stimulated" mostly appears to be a function of when I'm drinking it rather than a function of coffee volume consumed. Late afternoon seems like the time when caffeine is most effective (or maybe a little too effective, depending upon the level of focus the day requires). A few recent papers and at least one science blogger explain this effect in terms of cortisol levels and the circadian rhythm. See also: the concept of the zeitgeber.

I'm also reducing the level of anonymity for this blog, as you may have already noticed. This trend will continue. The person writing this is the same person who does everything else that person does every day.

Monday, April 07, 2014

Distant roses and yoga poses

Today I read about Sarcopoterium spinosum, a species of flowering plant common to the Mediterranean and Middle East. It's not an especially interesting plant - at least not to me. I'm not entirely certain why I focused on it. Perhaps it was the idea that this plant is quite common in some areas yet poorly studies. Some folks seem to think it's a potential source for an antidiabetic agent (but that's in the Journal of Ethnopharmacology; I neither have access to this journal nor can read the linked abstract without raising an eyebrow to the point of strain).

Unrelated: I'm been trying to take moments to be present lately, i.e. stopping to breathe and focus. I'm skeptical of most things which would look ideal on pamphlets for yoga classes but hey, meditating is really nice and much-maligned. Most notably, it pays to notice what's going on around you or perhaps even because of you. It's nice to notice when you're happy, too. Somehow that often appears to be treated as a separate idea (noticing you're happy, that is).

Dang, that guy is quotable.

Friday, April 04, 2014

I was trying to batch-rotate a whole bunch of photos today in Ubuntu 12.04 (it's 2014; maybe it's time to try out Trusty Tahr) but my usual solutions just weren't cutting it. Shotwell doesn't even have batch options, as far as I can tell. I've used digiKam before for batch editing but it doesn't even appear to import files anymore. Luckily, I found the Nautilus Image Converter. It can rotate and resize images right from the file manager, just like in, um, most recent versions of Windows. Anyway, problem solved.

Thursday, April 03, 2014

Music for today is Solyaris - Achieving Sky True through Vertigo Vortexes.

Anonymity, the Web, and whether people still use the term "the Web" in conversation

My Wikipedia username used to be a fiction. I changed it recently: it was time to cease being anonymous in that forum.* I didn't do it because of other Wikipedia users. If anything, I've always been a bit worried about the prospect that today's editor, blindingly-furious with me for my edits to their favorite page but completely incapable of actually communicating with anyone about it, will be tomorrow's supervisor. Academics can be tricky that way. By that same token, however, I changed my Wikipedia name so people begin to see my name attached to their favorite subjects.

I spend a fair amount of time writing about microbiology, whether it's online or for some kind of eventual peer-reviewed research**. The chances still seem pretty slim that academics will ever take Wikipedia seriously (and that's fine! It's just an encyclopedia, right?). Many of these researchers are unfortunately intent on citing Wikipedia in their articles. It's still pretty neat to think that the first words someone may read after googling the name of some bacterial species may be mine. Er, partially mine.

Maybe that whole shared-content aspect is the real stumbling block. Academics love the peer review process because it lets People Who Know What They're Doing evaluate People Who Are Doing Something Similar Yet Different. The process works just fine when it's partially anonymous; most people are kind in their everyday interactions but have a license to be critical when their identity is hidden. Pseudo-anonymity doesn't provide the same kind of protective shroud. It's really more of a dissociation effect in which a distinct aspect of one's personality is brought to the forefront.

I'm certainly not an expert in the field of anonymity. It doesn't matter because this is a place where I can write about things I may know very little about. I can do that more easily because my Real Name doesn't have to be responsible for it.  


*Not this one, though! If you're reading this, you already know who I am. It's the internet equivalent of wearing a false mustache.***

**That's all online now, too. I've done much more writing under a pseudonym than under my Real Name if you don't count classwork.

***The meatspace equivalent of wearing a false mustache on the internet is probably something like "breathing" or "eating lunch" or "sinking into a creative morass."

Wednesday, April 02, 2014

Today's paper is: 

The shortest-possible summary: Mycobacteria like M. tuberculosis have cell walls but they have nice, thick lipid membranes, too. All these layers add up to protection from antimicrobial compounds: the bacteria won't get killed if antimicrobials can't even get inside them.

Tuesday, April 01, 2014

Today, I read the following paper: 

This one is really just about a fairly simple method to assay protein-protein interactions. It's an advertisement: the method was developed commercially by ChromoTek Gmbh and is sold as a kit. The basic idea is that proteins are expressed in Baby Hamster Kidney* cells as GFP and RFP fusions. The GFP-fusion baits also have LacI, so they'll end up bound to lac in the nucleus no matter what happens. The RFP-fusion preys should just float around the cell unless they interact with the baits, at which point they'll also be present at the same spot as the GFP. The process is reversible, so an interaction can be observed in real-time as it is disrupted, i.e. by some variety of specific protein binding inhibitor.

I probably won't use this method much myself, but it's quite simple and looks like it could be convenient for screening potentially therapeutic small molecules or peptides. It would be nice to see a paper published about the method's applications by groups other than ChromoTek employees.


*There's something about the word "baby" that makes it difficult to take seriously. This is an established cell line and has been in use since the 60's.