Monday (19 November, 2007) saw the start of the two-day CISB07 conference, this year's internal conference for the 6 Centres for Integrated Systems Biology (for more information, see http://www.cisbs.org). There were also a number of invited keynote speakers. As I am unsure of the private/public status of many of the talks, I will just present some notes on those talks that I know are in the public domain.
CISB07 is being held at the Centre for Life in Newcastle. The Newcastle University CISB (CISBAN) is the host of the meeting, and I was one of about 8 people from CISBAN that helped organize it. It seems to have come together really well – it looked like everyone had a good time today. The Centre for Life is a science museum mainly for kids, but it has conferencing facilities too. I have to say that the people running the CfL today were very helpful and friendly. I'd definitely recommend the venue. After the afternoon session we had drinks on the mezzanine level of the museum, where many of the interactive (ok, and meant for children) games were. We had free reign on these games (reaction time games, soccer goal-scoring games, etc) and it was really fantastic to see all these scientists playing them with what can truly only be described as glee. Further, the CfL had their Motion Ride on for us – it's like a small cinema that plays a movie that's about 5 minutes long, and all of the chairs are on their own hydraulic pumps, and the chairs move in time with the movie. The theme tonight was "Dracula's Haunted House", and I have to admit I went on it – twice. Then it was in for dinner. However, it is the games that were really fun!
I gave a talk on SyMBA today as well, which I won't go into any more other than to say I'll put up the slides on the SourceForge site later this week, when I get a minute. As I can't be sure of the private/public status of any of the talks other than mine and the Data Integration Keynote (the keynote for the session I was in), I'll just pop up my notes from that keynote, from Dr. Chris Taylor of the EBI. It should be an interesting day tomorrow! Other than Chris' talk (notes below), I found that Michael Wilson's (from CPIB) and Mark Sansom's talks were also some of the more interesting talks of the day.
Standards development: A Two-Way Street (or, I Saw Six CISBs Come Sailing By)
(Chris Taylor, Keynote, Session 2)
There are three big omics standards bodies in biological science: MGED,
PSI, MSI. HUPO's parent organization is PSI, which deals with molecular
interactions, post-translational modifications, mass spec,
separations, gels, etc. They also have a number of formal groups e.g.
Steering Committee, working groups, and a document release process.
Many community formats are already built on FuGE-OM (GelML, mzML,
analysisXML, etc). These standards provide increased efficiency
(methods remain properly associated with the results generated, no need
to repeatedly construct sets of contextualizing information, for
industry specifically (in the light of 21 CFR Part 11)), enhanced
confidence in data (enables fully-informed assessment of results,
supports assessment of results that may have been generated months ago,
facilitates better-informed comparisons of data sets, supports the
discovery of sources of systematic or random error by correlating errors
with metadata features such as the date of the operator concerned,
follow-through with experiments performed), added value & tool
development (re-using existing data sets for a purpose significantly
different to that for which the data were generated, building aggregate
data sets containing similar data from different sources, integrating
data from different domains, design requirements become both explicit
Generic features of experiments; technologically and
biologically-delineated sections as well. But you can't ever really
carve these things up this way all the time, as technologies are
cross-biological community, and vice versa. The difficulty also comes
from trying to agree what a single term is from a mixed-community group
of people. This is why they didn't use experiment in their structure.
Instead, they use an Investigation-Study-Assay (ISA) structure. The ISA
structure starts out generic, and then you can extend it for your
community. RSBI (of which CISBAN is a part) includes MIBBI, FuGE, and
OBI. MI checklists are usually developed independently, which means
they're usually partially redundant. That's where MIBBI comes in. Where
they do overlap, they may cut things up differently or word things
differently. Which MI should you use? There could be more than one that
The CISBs are a prime source of researchers with a cross-domain
view. MIBBI has already established contact with all six CISBs, and is waiting for tech and research summaries from them. Overall, what he
wants is more interaction between systems biologists and
It was a humorous and informative talk, and I really think he got the point about standards and common formats across to the group.