Modeling:Mac
From QuB
| Prev: Modeling:MIP | Outline | Next: Modeling:Mac LiveFit |
Mac (Macroscopic Rate Optimizer) optimizes the rate constants (with error limits) of a user-defined kinetic model using macroscopic recordings (ensemble current, from hundreds or thousands of channels).
The parameters of the model may be constrained according to detailed balance, fixed rate constants, and proportionality of rate constants. Mac can optimize a stimulus-dependent model using data with a time-varying stimulus (recorded on its own A/D channel). Multiple files, derived from data obtained with different stimuli can be analysed simlutaneously to estimate the voltage or concentration dependence (with error estimates) of all rate constants. Mac can analyze channels whose single-opening current is a function of (changing) voltage.
Mac can used fixed initial (entry) probabilities (entered in State Properties), equilibrium entry probabilities (from the latest rate constants), or conditioning equilibrium: if for example you held voltage low, then flipped it high and started recording, Mac can initialize the state vector from equilibrium at the low voltage.
The details are explained in Dr. Lorin Milescu's thesis.
Properties
| Data channel index | which A/D channel contains the data, typically 0 |
|---|---|
| Quiet Output | print dramatically less info to the Report window |
| Data source | |
| Ligand, Voltage, ... | experimental conditions as needed by the model A variable can have a "Value" or take its changing value from an A/D channel. The "Conditioning" value is used to calculate conditioning equilibrium probability. |
| Add/Delete/Presets | (ignore these) add/delete experimental variables, and load/save all variables |
| Use | (column when Data source is File list) whether each file will be part of the file list |
| Pre-process data | none: use raw data as displayed: use the same baseline correction and filtering as the Data display as such: apply the specified filter and baseline |
| LL conv | stop if LL increases by less than this much |
| Grad conv | stop if all gradients are less than this much |
| Max iter | at most how many times to repeat (calculate LL and gradient, modify parameters) |
| Max step | how much to change parameters each iteration. 1.0 is the natural step size; smaller numbers may be more reliable for sensitive models, but will converge more slowly. |
| Restart once | if it stops after Max iterations, restart it (works better than increasing Max iterations) |
| Estimate channel count | |
| Estimate cc per file | |
| Amp(open) | |
| Var(open) | |
| LL (else SS) | |
| Min k | |
| Max k | |
| Initial probabilities | Fixed: use the "starting prob" from State Properties, normalized to sum to 1.0 Equilibrium: start from equilibrium, using the initial or constant conditions and latest rate constants Conditioning Eq: start from equilibrium, using the "conditioning" value of the stimulus. |
| Run mode | optimize (maximize LL) or check (compute LL with current parameters) |
| Batch (segments) | Together: Runs MIP with all the data at once, summing LL and generating one final model In groups of: In groups of: |
| Max batch | |
| Identical segs | |
| MUX files | |
| Propagate pars | |
| Fit data | Whole means nothing is skipped from the source data.
List means only the points included in the fit list will contribute to the LL, yet the probabilities will be calculated over the whole data source. If Exclude is checked, those data points included in the exclude list will not contribute to LL. Both can be used - sometimes it is easier to exclude than to include, sometimes the opposite is true. I use these options all the time. When fitting over multiple files, make sure the include/exclude lists have the same name in all files. |
| Skip (ms) | |
| Show fitting curve | Avg: Std: Res: |
| Skip slow changes | Breaks data into chunks with std < Max std, processing only one averaged datapoint per chunk |
| Max std | |
| I = f(Voltage) | |
| Vrev (mV) | |
| Gleak (?) | |
| Vleak (mV) | |
| Presets |
Batch Processing
The first "In groups of" option allows you to enter one number (n), using the up/down control. The other "In groups of" allows you to enter a sequence of numbers (n, m, ...), separated by commas or spaces etc. Then, supposing the data source is a list a many segments, Mac will fit them separately in groups of n. When second option is checked (n, m, ...), Mac will first fit in groups of n, then in groups of m etc. This is useful for statistics, when, for example, you want to see how the precision of the estimates depends on how many segments are fitted together (i.e., n, m etc). The reason we should keep both options (the second includes the first) is because it allows one to switch between one strategy and the other without having to remember what they were. The "together" option fits all the segments.
Results
A ton of relevant info is shown in the Report window. (explain please)
| Prev: Modeling:MIP | Outline | Next: Modeling:Mac LiveFit |
