Newcastle University/1 May 2008

Back

Meeting minutes
Minute taker: Nina Nielsen

Date: 01/05/08 Place: CT. 601 Time: 15.30-17.00

Attendees: 5/10 (Neil, Matt and advisors)

Supervisors:

Prof. Neil Wipat (anil.wipat@newcastle.ac.uk), Dr Jennifer Hallinan (j.s.hallinan@newcastle.ac.uk), Dr Matthew Pocock (matthew.pocock@newcastle.ac.uk),

Morgan Taschuk (mtaschuk@gmail.com)

Advisors:

Mike Cooling (m.cooling@auckland.ac.uk),

James Lawson (j.lawson@auckland.ac.uk),

Jan-WillemVeening

Team Members:

Nina Nielsen (nina.nielsen-dzumhur@newcastle.ac.uk), Mark Wappett (mark.wappett@newcastle.ac.uk),

Megan Aylward (megan.aylward@newcastle.ac.uk)

Meeting Content James no longer part of the team.

Quorum Sensing Presentation Mark explains attempt to find a quorum sensing system specific to Staph. Aureus. Discussion of Agr two component system Receptors of the pheromone are specific. Could this be a target specific to MRSA? Jen mentions the detection of one protein will not be enough for this project. But could be used in collaboration with other methods. Jen asks for the presentation and references to go on the wiki.

Individual project action plans Change meeting start time to 2.30 Then have half hour individual meetings with the supervisors (after an hour long group meeting).

Megan’s Actions

•	Morgan suggests start off storing the parts data in Microsoft Access. •	Jen suggests that since we are looking at transcriptional network there is only really a small collection of parts to consider: Promoters, RBS, coding sequences, terminators. Can get away with just looking at data for these parts.

Marks Action Plan

•	Look at a possible evolutionary algorithm •	Doesn’t have to be use real data. •	Just take random values and put them in optimal order. •	Look at Jen’s papers.

Nina Action Plan Research entity relational diagrams and normalization. Think about what types of you want to get out of it e.g. Kinetic parameters. Also just focus on the four parts mentioned above Expand later.

Jen mentions that would be brilliant if we could lay down some foundations for next year. Looking for biology undergrads - Mark mentions a mate.

Neural Networks

Jen gives neural networks presentation. Nonlinear relationship and can take in a range of inputs. So more flexible and powerful No one has tried to do it with bacteria before. Arose pretty soon after computers became mainstream. Bad at image analysis Computers tend to be good at what humans bad at. Humans are non linear but see patterns/relationships when not even there

Idea arose to use functionality of brain in a computer to make it less clinical

Modeled to a brain cell. Computers could never come close to complexity of human brain. Everytime a connection between two neurons was used, it strengthened. Can learn the weights that can take any arbitrary function Lots of different neural networks. Multi Layer perceptron is most commonly used. Have input nodes to represent level of peptides

1.	which constraints the parts have as well (size, sequence, optimal spaces, codon usage, complexity), as well as their behaviour.

2.	Turn this insilico model into a sequence and put dna back to biobrick repository.

Jan and matt talks about the methods to visualize the results by making key genes fluoresce. Use time lapse microscopy, flow cytometry. Orange and blue don’t work so well so limited to two or three colours only.

Next Meeting: 03/04/08 at 15.30 CT 701.