kDalton

kDalton

The kDalton program directly fits sedimentation equilibrium (SE) scans using non-linear least-squares analysis, providing global analysis of multiple loading concentrations and/or rotor speeds (up to 18 total data sets). It can analyze ideal or non-ideal self-association (with up to 2 different oligomers), and also provides 15 assembly models for mixed associations (hetero-association).

Program history

kDalton was originally developed in 1992-1993 while John Philo was in the Protein Chemistry group at Amgen. At that time there was no software available on personal computers to directly fit raw SE data, and in particular nothing that could handle mixed associations. In 1994 John developed a new approach using "soft" constraints to impose approximate mass conservation on the fitting solutions, and thereby exclude false, physically-unrealistic fitting solutions that otherwise plague least-squares fitting of mixed associations (see reference 5).

After John left Amgen in 1998 he continued to use, improve and develop the program, but for many years he could not distribute it because it was still considered to be a significant proprietary asset for Amgen. Finally in 2020 John ported the code to a newer compiler so that it could be run under Windows 10, added new features and a comprehensive Help file, and released the program for public use for the first time.

Program features

Globally fits up to 18 total SE data sets (different loading concentrations or rotor speeds). Those raw data sets can be at multiple wavelengths, and in centerpieces with either 3 or 12 mm pathlength.

True local Help file to give context-sensitive Help for each dialog box or control, with a searchable index.

Fitting results can be output in the units preferred by the user:

	true molar mass, buoyant molar mass, or reduced apparent molar mass (s)
	molar concentration, weight concentration, or signal units

Available reversible association models:

	monomer « dimer (ideal or non-ideal)
	monomer « N-mer (ideal or non-ideal)
	monomer « dimer « tetramer (ideal or non-ideal)
	monomer « N-mer « M-mer (ideal or non-ideal)
	monomer « N-mer + inactive monomer (ideal only)

Available mixed (hetero) association models (two types of monomer, A and B, all models ideal only):

	two non-interacting monomers (mixture of A and B)
	A + B « AB
	A + 2B « AB + B « AB₂, one type of site (both binding sites on A have the same binding affinity)
	A + 2B « AB + B « AB₂, two independent sites (the two binding sites on A have different affinities)
	A + 2B « AB + B « AB₂, two equivalent, cooperative sites
	A + 2B « AB + B « AB₂, two inequivalent, cooperative sites
	A + 2B « AB + B « AB₂: 2B « B₂, one type of site
	A + 2B « AB + B « AB₂: 2B « B₂, two independent sites
	2A + 2B « 2AB « ABBA
	2A + 2B « ... « ABBA, no B dimerization, K_dim(1) = K_dim(2)
	2A + 2B « ... « ABBA, no B dimerization, K_dim(1) ¹ K_dim(2)
	2A + 2B « ... « ABBA, K_dim(0) = K_dim(1) , K_dim(1) ¹ K_dim(2)
	2A + 2B « ... « ABBA, K_dim(0) ¹ K_dim(1) , K_dim(1) = K_dim(2)
	2A + 2B « ... « ABBA, K_dim(0) ¹ K_dim(1) ¹ K_dim(2)
	2A + 2B « ... « ABBA: 2B « B₂, 2 B binding sites, K1 = K2
	2A + 2B « ... « ABBA: 2B « B₂, 2 B binding sites, K1 ¹ K2

KDALTON will generate plots of the apparent molar mass vs. concentration. Such plots can be used to:

Estimate the maximum stoichiometry reached for self-association or mixed associations
Test whether the interactions are fully reversible
Test for significant solution non-ideality

Non-ideality is handled by iteratively solving for the monomer and oligomer concentrations at each radius, rather than using the actual signal for each experimental data point to estimate the total protein concentration for calculating the non-ideality effects, as was done in NONLIN. That shortcut used in NONLIN actually introduces experimental noise into the theoretical fitting function, and creates significant problems when the non-ideality is strong.

The mid-point of the fitted region is used as the radial reference position, rather than the first data point (as done in NONLIN and most other SE analysis programs). This choice reduces the cross-correlation between the parameters affecting the molar mass and the reference concentrations.

KDALTON will calculate and display the concentration and fraction of individual species averaged across the portion of the centrifuge cell actually included in the fit, and also compares the actual total signal across that region with the theoretical total signal based on the fit. Such information indicates whether all potential species were actually significantly populated in your experiment, and can guide designing an improved experiment.

Since the zero offset for absorbance data is typically independent of rotor speed, KDALTON optionally can constrain the offset values to be the same for a particular channel at different rotor speeds, reducing the number of fitting parameters.

Baseline scans can be subtracted from the raw data scans, which can eliminate the need to fit a zero offset and/or reduce the impact of scratches or dirt on the windows

If a baseline scan was recorded at a rotor speed different from that for a scan being fitted, you can optionally shift the baseline radially to try to compensate for the difference in rotor expansion

After a fit is completed, the program will generate a complete written report which details what data were fitted, what fitting model was used, and all parameters and user choices that affect the fitted results. That report can also optionally include either a residual plot or an overlay plot of the raw data and fitted curves.

Fits can be saved to disk and then later restored from disk. The saved file stores both the raw data files and all parameters and user choices that affect the fitted results.

KDALTON includes a calculator tool that helps you choose the correct rotor speeds for your experiment.

In addition to residual plots and overlays of raw data and fitted curves, KDALTON will also generate plots of species concentrations or species fraction vs. radius.

All generated graphs can be saved to a disk file so they can be exactly reproduced at a later time.

System Requirements

kDalton runs under Windows XP, Vista, 7, 8, or 10. Use via Windows dual-boot configurations on MacIntosh systems should work fine, but is not officially supported or guaranteed.