The BioBricks Foundation:Standards/Technical/Exchange/Application Scenarios

From OpenWetWare
Jump to navigationJump to search
  • Raik's Scenarios
    • Physical Part Exchange: Lab A sends a collection of samples to Lab B
      • What information needs to be exchanged along with it? (test for sequence, feature management)
    • Assembly Planning: Alex knows exactly what parts he wants to put together and needs to find the appropriate building blocks with matching assembly standard in a collection of basic and composite parts. (test for part / sup-part, format management) -- can perhaps be collapsed into Scenario (1)
    • Device Design: Bob has a device layout and needs to find suitable parts to implement it. He looks within JBEI or MIT or local registry. (test for part categorization, high-level information)
    • Characterization: Alex measures the performance of 10 promoters published by a JBEI team. He wants to associate experimental data to the parts hosted on JBEI.
    • Simulation: Deepak wants to simulate Bob's design using Alex' experimental data.
  • Tim's Scenario
  1. A plasmid Plas1 is created by assembling Part1, Part2, Part3.
  2. Plas1 is transformed into genotypically different Strain1 and Strain2.
  3. Experiment is performed on Strain1 and Strain2 harboring Plas1.
  4. One particular instance of Strain1+Plas1 (Ex1) displays useful behavior, but not others.
  5. Investigation of Ex1 reveal that its plasmid has a mutation in Part2. Call this Plas1' with modification Part2'.
  6. Extract Plas1' and retransform into fresh S1, and repeat the experiment.
  7. Repeat experiment fails to reproduce the result of Ex1.
  8. Further investigation reveals that Ex1 has a mutated strain S1'. Which means Ex1 is actually S1' with P1'.

I want to exchange the relevant data about this experiment with someone. At least the information on S1' and P1' has to be transmitted. It would be desirable to have info in S1' and P1' to "point to" their parents.

  • Deepak: Essentially we need a way to store homologous sequences as well as the phenotype (long description?) for each. There might be an "original" sequence in the set of homologs -- "original" from as experimental point of view, not evolutionary.
  • Raik 14:32, 13 August 2009 (EDT): Quite a sophisticated example Tim ;). This case could be covered by part inheritance (my suggestion in the data model diagram, not yet documented). The mutated Part2' could be created as a 'child part' of the original Part2. Only the differences would need to be annotated. (What is not covered by the current data model is the relation between different strains and plasmid backbones. We could revive some old PoBol idea and declare vector and strain/cell as sub-classes of "part". The inheritance would then also apply for them -- it would complicate our class hierarchy though.)
  • Cesar's Overall Architecture Diagram