TMT To Do List

Jacobian Bugs

1. Statistics from parameter estimation fails if there are more than 254 data points.
   • john can't duplicate
2. When running estimation from script, you can't specify which parameter to estimate (parameter number can't be a variable), wheras when running estimation block directly, parameter number can be a variable.
   • not a bug
3. Parameter sensitivities are not correct when the sequence section has a continue for 0 statement.
   • problem identified - will fix it
4. Jacobian greatly slows down after ~80 simulations are run from a single script.
   • John can't duplicate
5. Inconsistency with parameter variance in statistics of estimation report.

Experimental

• Get reporter working in glass bottom plate
• Get reporter working in Stimulator
• Check our microscope camera against molsci's (theirs is Micromax 512BFT)
• Try using poly-Lys instead of ConA

Papers

Stimulator paper

• Femlabs model of flow in chip. Force on cells.
• In Ste2 knockout and Ste2WT cells, see how long aF-fluorescein (David Drubin) takes to wash in and out of cell wall.
• Use different pH to dissociate pheromone faster?
• Hit cells with short (1-10s) dose of pheromone at different doses. What fraction at what concentrations respond and shmoo or express Pprm1-YFP?
• talk to Jeremy

Time-Dependent Stimulation paper

• Basis/core: identify time varying input timecourses that increase parameter sensitivity
• Perform same number of expts with step inputs and with time-varying input -> fit model in each case and compare parameter uncertainty
• Systematically look at different mutants (aF/Ste2, Sst2 deletion/overexpression, Ste2 overexpression, ...), and use of different potential outputs (if we could directly observe state of G protein then...., if we could measure nucleotide hydrolysis rate...).
• Examine full pathway and other parts of pathway (MAPK cascade) using potential inputs (Ste11 kinase inhibitor, Fus3 kinase inhibitor), and outputs (Fus3 phosphorylation, Dig2/Ste12 FRET, gene expression).