All the programmes except K2D rely upon reference datasets
of proteins, from which a set of basis spectra will be selected for the
analysis. CONTIN SELCON3 and CDSSTR offer a choice of reference database
which should be chosen in accordance with the range of input data. It should
also be noted that the choice of reference database affects the analysis
results particularly if there is mixed or high beta sheet content. The
reference set that represents the characteristics of the protein of interest
is likely to give the most accurate result.
| Reference Set | Wavelength range | No of Proteins | Author |
| Set 1 | 178 - 260 | 29 Proteins | C W Johnson |
| Set 2 | 178 - 260 | 22 Proteins | Johnson |
| Set 3 | 185 - 240 | 37 Proteins | 29 Johson
3 Venyaminov 5 Pancoska&Keiderling |
| Set 4 | 190 - 240 | 43 Proteins | 32 from SELCON3
6 Provencher&Glockner 5 Pancoska&Keiderling |
| Set 5 | 178 - 260 | 17 proteins | C Johnson |
| Set 6 | 185 - 240 | 42 Proteins | 32 from SELCON3
5 Pancoska & Kiederling 5 denatured |
| Set 7 | 190 - 240 | 48 Proteins | 32 from SELCON3
6 from Provecher & Glockner 5 from Pancoska & Keiderling 5 denatured |
| Set 8 | 178 - 260 | 43 proteins | a subset from Johnson |