Title: User Interface Paradigms for
1User Interface Paradigms for Describing Pathway
Models Cliff ShafferDepartment of Computer
ScienceVirginia Tech
2Fundamental Questions
- Who are the users?
- What are the tasks?
- What are the key bottlenecks?
3Who are the Users?
4Who are the Users?
- Modelers
- Expert users
- Observation Not biologists
- Chemists, Physicists, Engineers
5What is the Task?
- Primary Tasks
- Preparing a model for the purpose of simulation
- Figuring out the model
- Verifying the model
6What is the Task?
- Primary Tasks
- Preparing a model for the purpose of simulation
- Figuring out the model
- Verifying the model
- Other tasks are secondary
- Describing the model to others
- Exchanging models
- Teaching
7What is the Task?
- Primary Tasks
- Preparing a model for the purpose of simulation
- Figuring out the model
- Verifying the model
- Other tasks are secondary
- Describing the model to others
- Exchanging models
- Teaching
- Fundamental Problem Modelers are at the limit
8Modelers Lifecycle
Data Notebook
Wiring Diagram
Differential Equations
Parameter Values
Simulation
Analysis
Comparator
Data Notebook
9The Cell (Modeler) Cycle
- Outer Loop
- Define Reaction Equations
- Inner Loop
- Adjust parameters, initial conditions
10Tysons Frog Egg Model
11(No Transcript)
12Simulation of the budding yeast cell cycle
mass
Sic1
Cln2
Clb2
Cdh1
Cdc20
Time (min)
13Designing an Interface
14Designing an Interface
- Observation Rational human beings can have
widely divergent opinions on what interface
paradigm makes them productive.
15Designing an Interface
- Observation Rational human beings can have
widely divergent opinions on what interface
paradigm makes them productive. - Ex LaTeX vs. MS Word
16Designing an Interface
- Observation Rational human beings can have
widely divergent opinions on what interface
paradigm makes them productive. - Ex LaTeX vs. MS Word
- Observation Pathway modeling is a creative
process.
17Designing an Interface
- Observation Rational human beings can have
widely divergent opinions on what interface
paradigm makes them productive. - Ex LaTeX vs. MS Word
- Observation Pathway modeling is a creative
process. - Flow state
18What is Pathway Model Building Like?
19What is Pathway Model Building Like?
20What is Pathway Model Building Like?
- Photography?
- Programming?
21What is Pathway Model Building Like?
- Photography?
- Programming?
- Writing?
22What is Pathway Model Building Like?
- Photography?
- Programming?
- Writing?
- Drawing/Painting?
23What is Pathway Model Building Like?
- Photography?
- Programming?
- Writing?
- Drawing/Painting?
- Bridge designing?
24Existing Paradigms
- Graphical
- Virtual Cell, BSP, JDesigner
- Wizards
- Gepasi
- Equations or Scripts
- JigCell, Jarnac
- No implementation is pure
25BSP
26JigCell
Tysons Frog Egg Model
27A Characteristic of our Models
- We cant get away from the fact that there is
mathematics to describe - Parameter/initial condition values
- Rate laws
28How to Compare Interfaces
- Time required to complete task
- Errors committed during task
- Learning curve
29Application-Specific Issues
- Unavoidable text
- Total keystrokes required
- Density of information
- Support for abstraction
- Cognitive match for user
- Ability to grasp the overall structure
30Graphical vs. Equation
- Pros
- Better intuitive grasp of overall model
- Cons
- Low information density
- Poor support for mathematical details
31A Usage Scenario
- Modeler is struggling with setting the
fundamental structure of the model - Graphical sketch
- Modeler needs to fill in the details
- Fill in skeleton on spreadsheet
- Modeler is in verification mode
- Alternate simulation runs with tweaking the
spreadsheet - Modeler wants to explain to others
- Show a diagram