Quick start
Open your own ITC dataset, process it, fit a one-set-of-sites model, review the result, save the project, and export a figure.
Last verified:
This walkthrough takes a compatible file from measured heats to a fitted result. A One-Set-Of-Sites model is a starting example when the binding sites can reasonably be treated alike; choose another model if the chemistry calls for it. The steps show how to use the application, not how to decide which scientific model is true.
Before you begin
Use a compatible raw thermogram, integrated-heats file, Origin project, or .ftxtc project; see Supported input formats. Integrated heats and Origin projects without a usable time/power trace skip thermogram processing.
1. Open your data
Launch FT-ITC Analysis and choose Open File... on the welcome screen, choose File > Open..., or drag compatible files into the application window. Select your file and choose Open.
Select the imported experiment in the data list and open Overview. See Workspace for selection, Active experiments, and list ordering.
2. Edit experiment details
Open Details... for the selected experiment. Concentrations, conditions, comments, and attributes can be edited here. A date read from the data file is trusted and read-only. A filesystem timestamp does not establish the experiment date; entering a replacement marks it as User provided. The date source is shown beside the date controls.
Tandem experiments keep their generated merge description under Tandem merge origin, separately from editable comments.
3. Process a raw thermogram
If the import contains a thermogram, open Process Data.
- The baseline estimates the signal the instrument would show between injection peaks. If that background drifts smoothly, try Polynomial. For a more irregular background, try Spline; for changes that differ across parts of the run, try Segmented. Keep the default integration settings initially.
- Inspect whether the baseline represents the signal between injections rather than the peaks.
- Inspect the start and end of every integration region. A region should include the injection response without extending unnecessarily into baseline noise.
These baseline alternatives are explained in Processing.
4. Fit a model
Open Analyze Data, choose Single experiment, and select One-Set-Of-Sites or a more appropriate model, depending on your system.
- Review the initial parameter values: affinity describes how strongly the substances bind, enthalpy describes the heat change associated with binding, and stoichiometry describes the binding ratio in the chosen model. Use physically plausible starting values.
- Choose an optimizer. Levenberg-Marquardt is efficient near a suitable solution; Nelder-Mead provides a derivative-free alternative when convergence is difficult.
- Optionally enable Weight by injection error when the integration uncertainties are meaningful for the dataset.
- Choose an uncertainty method. None skips parameter uncertainty; Bootstrap residuals refits simulated variations of the data; Leave-one-out checks how much the fit changes when an observation is omitted; Profile likelihood tests how far a parameter can move while the fit remains acceptable.
- Choose Run Fit.
Create analysis result determines whether a usable fit also creates a separate Analysis Result. Enable it to continue through the result workspace in the next step; otherwise, the fitted solution remains attached to the Experiment Data.
If the fit fails or reaches a limit, the possible causes are described under Fit availability and non-convergence.
5. Review the result
Inspect the fitted curve together with the residuals (observed heat minus predicted heat). Random-looking small residuals are more reassuring than a repeated pattern the model misses. When Create analysis result is enabled, select the resulting Analysis Result and check:
- the included experiment and injections;
- fitted parameter values and units;
- convergence status and RMSD;
- whether weighting and uncertainty estimation match your intention;
- parameter uncertainty or confidence intervals, when calculated;
- parameter correlations.
6. Save a portable project
Choose File > Save As... and save the project as .ftxtc. The project contains imported data, experiment details, processing state, fits, results, and available advanced-analysis output. Keep the original instrument files as source records even though the project is portable.
If autosave is enabled, it supplements normal saving; it is not a replacement for a named project file.
7. Export a figure
Return to the experiment and open Final Figure. Choose the elements you need, such as the thermogram, integrated heats, fitted curve, residuals, confidence band, or parameter information. Review axis ranges and labels, then use the figure export or print command provided by the view.
Use Analysis Result Exporter... when you need numerical result tables rather than a graphic. For injection-level heats, choose File > Export Integrated Peaks.... The same output is available through File > Export Data... by choosing Integrated Peaks.
The figure and table workflows continue in Figures and export.