Monday, June 28, 2010

Weeknote #7 (w/e 27/6/10)

While flicking through the June issue of the BBC's Focus magazine, I noticed that one of my research collaborators had received a nice mention from Ian Stewart at Warwick. He was asked to select three books on puzzles and games; Martin Gardner was the obvious first-choice author, and Winning Ways for your Mathematical Plays is a minor classic. Stewart's final choice was a book written by my collaborator at New York University, Dennis Shasha. In the column, Stewart describes Dr Ecco's Cyberpuzzles as "...a fantastic book if you want to spend some serious time solving puzzles and giving your brain a work-out."

Ian Stewart has been a significant influence on my career; as a popular science author, I've always been impressed by his writing, but he had a rather more direct effect on me back in the mid-1990s, when I was a graduate student at the University of Warwick. Ian very kindly wrote me a reference to attend the prestigious Complex Systems Summer School at the Santa Fe Institute, and the month I spent there was incredibly important in terms of shaping my personal ambitions and outlook on research.

Now, I'm fortunate in being able to collaborate with people of Dennis' calibre (see the previous note, below), and last week he very kindly sent me a copy of his latest book. Co-written with Cathy Lazere, Natural Computing is a profile of the frontiers of computer science, told through the stories of fourteen pioneers, such as Rodney Brooks, Ned Seeman and Paul Rothemund. I'll post a full review once I've finished it.

Monday, June 21, 2010

Weeknote #6 (w/e 20/6/10)

We (three colleagues and myself) were recently successful in obtaining funding from the NanoInfoBio project to test an idea that's been rattling around for a while. DNA hash pooling is a technique that Dennis Shasha developed, with some assistance from me, while I was visiting him. Dennis is an incredibly sharp and prolific Professor of Computer Science at the Courant Institute of New York University. He was the Series Editor for my first book, and we kept in touch since its publication. Justine, the little one and I visited Dennis while he was in Paris on sabbatical with his family, in the summer of 2007. While Tyler, Dennis and Karen's son, played American football, we walked round and round an athletics track on the edge of the city, knocking around our own particular problem.

The task of analysing large populations of mixed DNA strands is of particular relevance to the emerging field of metagenomics, which is concerned with understanding, in genetic terms, the vast complexity of the planet's biosphere. Methods for looking at environmental samples often require a lot of genetic sequencing; although new ways of doing this are constantly driving down the cost, it can still be expensive to sequence large populations, as well as time-consuming. Dennis and I developed a technique that combines computational analysis with simple rounds of laboratory steps, based on the computer science idea of hashing. The idea is to associate "labels" with individual sub-populations of genetic sequences, such that the number of different genomes with the same label is relatively low. In this way, each genome (or genomic fragment) is associated with its own "fingerprint", which we can then use to confirm its presence (or otherwise) in a sample. Our hope was that this technique would offer a cheap, quick and simple pre-processing step before any sequencing was required, thus reducing the cost and complexity of analysing a sample.

We finally published the theoretical paper last year, but have only just obtained the funding to actually test the idea in the lab. I floated the concept at one of the NIB brain-storming meetings, and it was picked up by a talented team of biologists (Trish Linton, Mike Dempsey and Robin Sen). We put together a proposal to NIB for a small amount of support (£25K), and we were fortunate enough to be one of three projects funded in the last round. The nine-month post-doctoral position is currently going through the MMU approval process, so watch this space if you're interested.

Monday, June 14, 2010

Weeknote #5 (w/e 13/6/10)

The focus of the past week has been on Getting Things Done. After what's been probably my busiest academic year so far, I finally decided that my workload was such that I required a rigourous approach to task management. I trawled around for methodologies that would allow me to organize a multitude of different jobs, whilst maximizing the time I could spend with my family. After reading about Getting Things Done (GTD) on Merlin Mann's well-respected 43 Folders blog, I decided to give it a go. There's a nice "getting started with GTD" article on 43 Folders, which summarises the approach thus:

  1. identify all the stuff in your life that isn’t in the right place (close all open loops)
  2. get rid of the stuff that isn’t yours or you don’t need right now
  3. create a right place that you trust and that supports your working style and values
  4. put your stuff in the right place, consistently
  5. do your stuff in a way that honors your time, your energy, and the context of any given moment
  6. iterate and refactor mercilessly

And that's it, really. Most of the week was spent on the first three steps (the creator of GTD recommends at least a couple of solid days), but the effort was well worth it. I started by taking the various slush piles, to-do lists and marked-up journals and papers in my home office, and merging them into one big "in" pile. I then had to do the same with my work and Gmail inboxes, extracting only the "open loops" (i.e., unfinished projects).

I had over 4,000 emails in my Gmail inbox, and working through the whole lot, deleting as I went, quickly lost its appeal. I therefore adopted a "tagging" approach; I created an "@action" tag in red, and then skimmed through my inbox, tagging anything that required an action on my part. Everything was then selected and archived (just "select all", answer "yes" when it asks you if you want to apply this to all conversations, and then hit "Archive"), leaving nothing in my inbox (for the first time in many years). I could then select only the tagged messages, which was much more manageable.

The end result of this physical and electronic clear-out was a car-full of paper to go to the recycling centre, a clean workspace (shown above) devoid of distracting piles of paper, and a fresh outlook on work. I'm already feeling the mental benefit, as I've been relieved of the self-inflicted stress brought on by my subconscious constantly asking "what am I currently not doing?" I've always been quite cynical in the past about "snake oil", management-driven "productivity" schemes, but I can honestly say that GTD is an eye-opener, and it actually seems to work.

I've managed to condense everything down to a list of just over forty "projects" (ranging from "Fix external hard drive" to "Write next book"), most of which have a discrete "next action" attached to them.

I'll be writing more about GTD in the coming weeks and months, as I learn more about the system and (hopefully) realise its potential.

Monday, June 07, 2010

Weeknote #4 (w/e 6/6/10)

I've spent the past week in Madrid, at the Universidad Politecnica. I was a Visiting Professor in the Faculty of Informatics, delivering a series of lectures on "molecular and cellular computing" to their Masters-level students.

In the past, some people have expressed an interest in the material, so I thought I'd make it available here. A lot of it is based on my book Theoretical and Experimental DNA Computation (Springer, 2005), although there's a lot of new material in the second half of the series.

The lectures are as follows (links to PDF versions of the slides):

Day 1: Molecular Computing

1. Introduction and historical motivation.

2. The first experiment.

3. Subsequent work.

Day 2: From in vitro to in vivo

1. Models, lab work, and the transition.

2. Laboratory implementations.

Day 3: Biological Engineering

1. Biological background.

2. Synthetic biology.

3. Synthetic Biology II.

Creative Commons License
Molecular and Cellular Computing course material by Martyn Amos is licensed under a Creative Commons Attribution-Non-Commercial-No Derivative Works 2.0 UK: England & Wales License.