Jaunty Jackalope

March 27, 2009 by Steve · Leave a Comment 

Ubuntu 9.04 – Jaunty Jackalope beta has hit the Canonicle servers.  Full release is due at the end of April.  I’m going to give the 32bit desktop edition a run out on my laptop over the weekend.

9.04 promises Open Office 3.01 and the new ext4 filesystem which is reporting faster boot up times.  No 2.6.29 kernel however as the development team will freeze the feature set before its release.  9.04 will ship with 2.6.28 stable.  Ubuntu 9.10 will be shipping with 2.6.30 or 2.6.31… or neither if we’re all credit crunched to oblivion :p

FEBS feedback

March 18, 2009 by Steve · Leave a Comment 

There wasn’t much opportunity to obtain internet access at this years FEBS 2009 conference.  We had a busy schedule of lectures from 8.30am to 10ish, followed by practical workshops, blackboard sessions, and discusson groups until lunch time.  We were given a break from 1-4pm to take in the local scenery, particularly the ski slopes after which we returned for lectures and discussion sessions until 6pm.  The evenings involved a meal until 7pm after which we returned for poster sessions and group discussions until 11pm.  This was a challenging schedule and crammed an enormous amount of science into each day.

I particularly enjoyed the talks from Roger Brent from Berkeley, CA and Ron Weiss of Princeton as these were particularly relevant to my work in synthetic biology.  The electrical engineering approach of Ron Weiss was very useful in conceptualising the design of synthetic gene circuitry and his work in mamalian cells was inspiring.  Roger Brent provided some practical methods of measuring outputs of gene circuits and some useful open source software tools and lab methods that will greatly benefit my project.

There was also some very good talks from Eda Klipp on the mathematical modelling of biological pathways, and combined with practical copasi sessions with Frank Bruggerman modelling MAPK signaling pathways.  There was a lot of practical worked examples combining wet lab experimentation with mathematical modelling that gave a real incite into the development of systems biology and the application of it’s approach.

Denis Noble popped in for a keynote lecture on his development of the heart model and provided a unique opportunity to gain an understanding of the processes he went through in building one of the pioneering projects in systems biology.  There was also a nice opportunity to meet with him and we received a signed copy of his book - the music of life.  It is always an inspiration to hear Denis Noble speak and I would recommend any scientist to take a look at his talks on youtube.

I will update more info as I get time, as it’s full steam ahead on my return to Manchester.

FEBS systems biology advanced lecture course 2009

March 4, 2009 by Steve · Leave a Comment 

febs2009

This year our DTC studentship will be attending the FEBS systems biology advanced lecture course in Alpbach, Austria.

I have booked in for the “physiological systems biology” and “modeling cell biology workshops”.  Hopefully both courses will provide valuable training for the dry component of my project.

I will be updating from Austria if I can get internet access.

choose your weapon

March 4, 2009 by Steve · Leave a Comment 

Faced with the task of creating a PhD dissertation over the next 3 years I briefly investigated alternatives to the standard Microsoft Office suite.  I had played with Lyx last year but I quickly found myself outgrowing the WYSIWYG interface as I learned to code.  I also required a reference manager as my literature review was comprising over 200 references and the easiest way to get these software packages to interact was to dump the crutches.

I have tried to stay away from Microsoft Office this time as I have spent many years wrestling with auto correct, auto formatting, and frustrating middle of the night sessions watching figures jump around pages and printed documents looking nothing like what is on the screen, not to mention the random events that will occur when I send it to another machine.  I have also moved over to Ubuntu 8.10 as my primary operating system and adopted open office.  Still, the prospect of writing a thousand page document in MS or open office didn’t feel practical.  I stuck with open office initially and needed to investigate a reference manager that would play with it.  The Universities recommendation was Endnote but this was out on my Linux box so I began with bibus as it can store your references in an sql database and I had visions of being able to keep my library on a web server.  Unfortunately I ran into problems when exporting the database – it wouldn’t, and it didn’t play very nicely with open office, requiring time to install pipelines to pass data between the two applications, and generally being unpleasant if I wanted to rapidly throw in reference on the fly.  Bibus also had problems finding references in pubmed and often couldn’t find them even with PUID numbers.  As with bibus, Endnote also appeared to require a break in the writing flow to input a reference.

I cut my losses with office suites and took the plunge into LaTeX proper with Tex Maker and kbib.  Kbib so far hasn’t failed to find a reference search I’ve thrown at it, and imports citations from the web and exports my library into various RIS, xml, and endnote formats ensuring some future compatibility/portability.  Tex Maker isn’t the best LaTeX editor and is still digital blasphemy for hardcore vim or Emacs users, but I like the help it gives me with the code and syntax.  Kile is better with in-line spell check, but texmaker gives me a native gnome application and I can do bulk spell checks periodically.  The code was a hurdle at first, but after a few pages I had grasped a set of basic formatting codes and LaTeX was allowing me to just write and not think about headings and margins, and where the heck figure 9 just went.  I also seemed to grasp the equation editor relatively quickly and found it fairly straight forward to use.  It isn’t as simple as the MS and open office equation editors, but allowed me much greater control.  The real power of LaTex for me has been the labelling and tagging capability.  I can just tell LaTex what’s a figure, equation, diagram, etc and carry on writing.  References are much the same, just inputting the library identifier and carry on with the text.  The compiler will then number and reference them automatically, which is an enormous time saver and doesn’t break the flow of writing.  Inserting a new figure in the middle of a 300 page document at 3am the night before submission is a horror I have experienced plenty of times before in Office, and manually curating a bibliography is impossible this time.  LaTeX isn’t all loved up for the user though.  Document compiling can be a daunting experience when a random $ sign can give you a completely incomprehensible error message at 2am requiring an hour of debugging.  This is where MS Office scores a few dozen points back.  LaTex GUI’s could definitely improve the bug hunting for “noobies”.  After 3 months though I am fairly happy now creating all of my documents in LaTeX and would recommend anybody creating anything greater than an essay to use it, as you can quite literally just write and let the compiler worry about everything else.  Happy days.

For our impending conference visit I attempted to create a scientific poster in LaTex.  Error.  I gave this a good go… honest.  However I found my LaTex Jedi skills to be somewhat under developed and there just wasn’t enough control over the layout without a serious time investment in learning some more advanced code to complete the task.  The University recommended MS PowerPoint, however an A0 document with images in Powerpoint is a nightmare without a quad core processor and as much ram as you can stuff inside the case.  Open Office also requires an enormous amount of time to load and save this kind of document, especially converting between MS office formats.  This time I turned to Scribus, an open source desktop publishing package that was designed by a LaTeX user also struggling to create scientific posters and struggling with the lack of available software. I hadn’t really used Scribus before but found it very easy to input text boxes and images, with fine control of positioning using a WYSIWYG interface.  Scribus also exports as PDF so you can export your documents and send them to the printers moderately confident of what will come back.  Scribus is also infinitely more sophisticated than powerpoint with support for scalable vector graphics and templates for almost any kind of document.  Happy days again.

My next challenges are finding some alternative graphing software to escape Excel and some training in Matlab.

blog rebooted

March 3, 2009 by Steve · Leave a Comment 

I have reactived my blog to keep an account of my work in systems biology, inspired by Professor Douglas Kell’s blog at the BBSRC.  I am also due to attend the FEBS systems biology conference in Austria next week and would like to record the event.

I will see where it goes…