There are three categories of limitations in the calculations implemented in Sednterp. The first category is limitations implied in choosing a method of interpretation. Sednterp implements certain general models for interpretation which are not always the best models for a certain experiment. This kind of limitation can be overcome by extending the principles used in Sednterp to include more sophisticated algorithms. In these cases Sednterp may still be useful to compute some parameters needed in these more sophisticated algorithms. Users who often use an algorithm not implemented in Sednterp should contact the programmers; portions of the source code are available to interested parties and suggestions for extensions of general use are welcome. For example, modeling proteins as ellipsoids of revolution is most applicable to the examination of monomeric, globular proteins, and is less useful in the detailed examination of oligomeric structures. To overcome this limitation, either more exact or more general models should be used. (ref 58) However, the use of such models requires that the experimenter already have some structural information available. In these cases, provided that sufficient material of high enough purity is available, there are other methods which can provide more detailed structural information. That is to say, the real power of analytical ultracentrifugation is to provide an initial, rough idea of the asymmetry or hydration of a protein. (Ref. 56 57) If other structural data exist, then difference sedimentation velocity methods can provide useful information concerning the changes in structure. (Ref. 9 10 11 12 32 56 57)
The second category is limitations fundamental to the use of predictive schemes. Many of the equations are derived assuming a two-component solution, even though experiments routinely are performed in buffers of three or more components. On theoretical grounds it is clear that this assumption is not correct. However, over a reasonable protein concentration range and when thorough dialysis has been performed against solvents of moderate pH (i.e. where the net-charge to molecular weight ratio of the protein is less than about 1:5000 mole/g) and moderate ionic strength (0.05-0.5 M), the sedimentation behavior of many proteins is described adequately. (Ref. 2 3) It is still important that the experimenter be aware that neglect of complexities resulting from multi-component solutions can lead to misinterpretation of data. (Ref. 3 13)
The third category is limitations based on the quick and easy interpolating functions used by Sednterp. Without compelling reason, none of the calculated or predicted values should be used as a substitute for experimental determinations. For example, the molecular weight calculated and should not be used as a substitute for the weight calculated from a known composition, unless there is a compelling reason such as post-translational modification of the protein. Likewise, the values of vbar, ρ and η computed by Sednterp should never be used as a substitute for experimentally determined values. For example, due to non-additive volume changes on mixing, densities calculated by Sednterp using equation 13 may be in serious error when two or more solutes are present in moderate concentrations. This same problem can arise for mixtures containing both salts and a high mole fraction of non-ionic components. In fact, Sednterp will compute densities and viscosities of buffers that could never be created. For every physical parameter that Sednterp interpolates or estimates an entry field for exact experimental data is included.