Newcastle University/8 May 2008

Back

Meeting Minutes

Minute taker = Mark Wappett

Action Plan for next week

Group •	Paragraph for external Wiki.

•	Make an executive decision on the pathogens to be detected, and draw up a rationale for their detection.

•	Begin Planning the Marketing strategy.

•	Become familiar with the structure of CellML.

Megan •	Build ER model

•	Come up with a list of parts

•	Write a ‘Back of envelope,’ description of those parts

•	Look at Simulator software

Mark

•	Look at neural network software

•	Enter dummy data to this

•	What does it need to be told

•	What needs to be defined by input and output parts

The meeting

Time: 14:30 – 17:00

Attended by: Neil Wipat – Supervisor

Jen Hallinan – Supervisor Matthew Pocock – Supervisor Jan-Willem – Associate

Morgan Taschuk – Project 5

Mark Wappett – Project 3 Megan Aylward – Project 1

Meeting Content

Mark Gives overview of achievements for the past week, and outlines what he and Megan Aylward have done. Subject turns onto MRSA, in particular

•	Could we just try to detect the presence of Staphylococcus/Streptococcus?

•	Look up Agr Quorum sensing regions + Blast. Perform alignment and search for any INDELS in the agr region.

•	Other possible targets – Staphylococcus epidermis, Clostridium difficile, Staphylococcus anthrasis

•	Possible rationale could be to get the system working on specific strains and then fine tune to conditions such as MRSA

•	Could the bug become a sort of gram positive pathogen buster?

Jen Asks whether or not simulator software has been looked at over the past week? It has not been, and therefore this must be carried out before the next meeting.

Subject turns to Cambridge, in particular:

•	Their entry – the bacterial internet

•	Might be coming to talk to us, and we may be able to steal some of their bricks

Jen and Neil mention Rea, the second year Microbiology student that will be in the Wet lab over the 10 week summer period

Subject turns to the Specification of Parts, in particular:

•	The use of Transcription factors and their promoters

•	Possibility of using a sporulation mutant

•	The potential use of sigma-factors instead of transcription factors – Jan-Willem

Neil •	Need to decide on the layers in the neural net

•	Numbers of nodes

•	Can sample from the genome + develop own promoter library

•	Can only pick things that we know will work

•	Consider possible cross talk between sigma factors

Possible parts + functions were then written on the board

•	LacI - Repressor

•	XylR

•	LacO – Operator for LacI

•	TetR

•	ssrA

•	ComK – activator that does not require phosphorylation to work

•	PcomG – Promoter

•	Pspank - Promoter

•	Fi29

Matt talks about control with Jan-Willem

•	Relative levels of activator/repressor will be important

•	Promoter strength

•	Repression strength

•	Leakiness/repressor stability

Jan-Willem suggests that we read the Van Oud enaarden, Grossman paper that characterises the Pspank promoter.

Neil recaps on the function of Sigma factors

•	Modify promoter binding

•	Act like a key on RNA polymerases

General subject turns to modulating strength of repression and promoters

•	Some genes can have four bacterial promoters

•	Ribosome binding sites may be good consistent parts for the neural network

Jan-Willem exits.

Neil reiterates earlier points for Megan’s benefit

•	Use range of peptides from range of Staph and/or Strep gram positive organisms

•	Weights in the potential neural network will be replaced by levels of affinity

Megan outlines her work for the past week

•	Constructed her database

•	Talks about problems with formats and problems with ontology’s

Matt turns the subject to competing projects:

•	Need to get our project information online and on the Wiki so that our idea does not get stolen

•	Need also to keep an eye on other projects and the state of their Wikis, so that we can modify our appropriately and keep up. Suggests 5 projects each to monitor

The subject then turns to Marketing material

•	Required to secure a decent amount of grant money

•	Web-site design needs to be considered

•	Internal and external wikis need to be presented well

•	Look at other groups ideas and steal them??

•	Could even do a promotional video!

Subject then turns to agendas and minutes, in particular:

•	Each week there must be an individual report given on progress

•	There must also be a group report

•	Each agenda must incorporate action points from previous weeks and how they were addressed must be delivered

•	Must have new points for the next week

•	Minutes must have action points on them

Morgan then talks about the need for interfaces and the problems she is having, in particular

•	Needs to know how everything is going to talk to each other

•	Currently parts are given in tree form e.g. promoters, and coding regions (such as GFP) and levels are given as low, medium or high

•	The EA data manager – levels of inputs and outputs must be known

•	A fitness function needs to be calculated and there are no simulation functions yet •	How do you pass a part around?

Meeting then turns to focus on Megan and the definition of her parts. A list is drawn up highlighting the parameters that each part must encompass:

Megan should now be able to come up with a preliminary list of parts and parameter and construct an entity relationship diagram from this

Meeting then turns to focus on Mark and the Evolutionary algorithm, points raised:

•	Layers of the network need to be defined

•	Space of models known as the constraints

•	Real data – Should be used to play with the neural network simulator to aid with understanding of neural nets

•	Conceptual understanding of a neural network. Will not be changing the order of parts, and some may well be distinct such as a set RBS.

•	However part mutations will be the changing of the parameters of the parts to create new instances and to see how the part model reacts

Mark should no be able to better understand the EA concepts, and plan the neural network in the coming week

Meeting is wrapped up with the action points listed at the top of this document.