Manual


Global Nonlinear Least-Squares Equilibrium Analysis:

Part 7: Model-Specific Information Previous: Fitting Information

The UltraScan global nonlinear least squares fitting program for equilibrium data provides a number of predefined models, each is discussed in further detail below. Each model has a global model number, which is used in default naming conventions for result files, Monte Carlo analyses and as a model descriptor across different methods. The model number is indicated in each description:

  1. 1-Component, Ideal: This model can be used for a pure, single component system that behaves hydrodynamically ideal. This model should be used first, if model fits well, the system is homogeneous and doesn't self-associate. The following parameters are fitted:

    The model number for this model is "0".

  2. 2-Component, Ideal, Noninteracting: This model can be used for for a system of two components that are not interacting, and behaving hydrodynamically ideal. After fitting this model, and a good fit was obtained, it is a good idea to compare the ratios of the integrals between scans run under different conditions.

    If the ratios are roughly unchanged from scan to scan (by an order of magnitude), there is a good chance that the system is truly noninteracting and two different components are in the system. If the ratios change in the direction of larger amounts for the larger component with increasing concentration and speed, the model should be fitted with a self-associating model, since the molecular weight average changes with concentration distribution, and a concentration-dependent self-association is most likely present. The following parameters are fitted:

    The model number for this model is "1".

  3. 3-Component, Ideal, Noninteracting: This model can be used for for a system of three components that are not interacting, and behaving hydrodynamically ideal. After fitting this model, and a good fit was obtained, it is a good idea to compare the ratios of the integrals between scans run under different conditions. If the ratios are roughly unchanged from scan to scan (by an order of magnitude), there is a good chance that the system is truly noninteracting and three different components are in the system.

    If the ratios change in the direction of larger amounts for the larger components with increasing concentration and speed, the model should be fitted with a self-associating model, since the molecular weight average changes with concentration distribution, and a concentration-dependent self-association is most likely present.

    Because this model has a large number of degrees of freedom, this model should only be used for cases where all other models fail to describe the system. Rarely do equilibrium scans contain enough information to quantitatively describe a three component system with all of its parameters floated. However, if you do know the molecular weight of each species, you can leave those parameters fixed and just fit the amplitudes and baseline. However, this would be a linear fit, and is more appropriately dealt with in the next model, the fixed molecular weight distribution, described below. The following parameters are fitted:

    The model number for this model is "2".

  4. Fixed Molecular Weight Distribution: In this model, the molecular weights are preset as an evenly divided distribution of molecular weights between some lower and upper molecular weight limit. This model can be used for any experiment, since it fits the experiment in an almost model-independent way. Several diagnostic plots are provided with this model to ascertain what nonlinear model may be most appropriate for fitting. Fitting with this model can prove helpful for cases where you need to distinguish between self-associating and noninteracting systems.

    A predetermined molecular weight distribution with the molecular weight parameter kept fixed is used to fit all scans independently with general least squares. Since no exponents are fitted, the fitting function can be considered a linear combination of nonlinear terms, which is linear in the coefficients that are fitted. The coefficients are nothing more than the amplitudes for each exponential term with a different fixed molecular weight. Use at least 3 different molecular weight terms to describe a model-independent system.

    If the residuals for this fit do not come out perfectly random and show systematic drift, the molecular weight distribution is not set up correctly. Perform a single component, ideal fit first, and repeat the fixed molecular weight distribution model by centering your molecular weight distribution around the molecular weight obtained by the single component fit.

    After the fit completes with satisfactory residuals, you can display the results in one of three ways:

    The plotting functions listed above are available for this model only and can be accessed from the fitting control panel.

    The following parameters are fitted:

    The following parameters are fixed:

    The model number for this model is "3".

  5. Monomer-Dimer Equilibrium: This model can be used to fit an ideal, reversibly self-associating monomer-dimer system. The following parameters are fitted:

    The model number for this model is "4".

  6. Monomer-Trimer Equilibrium: This model is identical to the monomer-dimer equilibrium model, except instead of a dimer the association is for a trimer.

    The model number for this model is "5".

  7. Monomer-Tetramer Equilibrium: This model is identical to the monomer-dimer equilibrium model, except instead of a dimer the association is for a tetramer.

    The model number for this model is "6".

  8. Monomer-Pentamer Equilibrium: This model is identical to the monomer-dimer equilibrium model, except instead of a dimer the association is for a pentamer.

    The model number for this model is "7".

  9. Monomer-Hexamer Equilibrium: This model is identical to the monomer-dimer equilibrium model, except instead of a dimer the association is for a hexamer.

    The model number for this model is "8".

  10. Monomer-Heptamer Equilibrium: This model is identical to the monomer-dimer equilibrium model, except instead of a dimer the association is for a heptamer.

    The model number for this model is "9".

  11. User-defined Monomer-Nmer Equilibrium: This model is identical to the monomer-dimer equilibrium model, except instead of a dimer the association is for a user-defined stoichiometry. You can define your desired stoichiometry in the monomer-Nmer stoichiometry selection panel.

    The model number for this model is "10".

  12. Monomer-Dimer-Trimer Equilibrium: This model can be used to fit an ideal, reversibly self-associating monomer-dimer-trimer system. The following parameters are fitted:

    The model number for this model is "11".

  13. Monomer-Dimer-Tetramer Equilibrium: This model can be used to fit an ideal, reversibly self-associating monomer-dimer-tetramer system. The following parameters are fitted:

    The model number for this model is "12".

  14. User-defined Monomer - N-mer - M-mer Equilibrium: This model can be used to fit an ideal, reversibly self-associating monomer - N-mer - M-mer system. You can define the stoichiometries for the N and M associations using the stoichiometry selection panel. The following parameters are fitted:

    The model number for this model is "13".

Previous: Fitting Information Model-Specific Information


www contact: Borries Demeler

This document is part of the UltraScan Software Documentation distribution.
Copyright © notice.

The latest version of this document can always be found at:

    http://www.ultrascan.uthscsa.edu

Last modified on January 12, 2003.