Abstract
Different approaches to normalization for spectroscopic data mining are discussed. A widely used approach to spectra normalization is to adjust all data to a region of known intensity. Statistical approaches to normalization are based on calculating some property of the spectrum that relates to the magnitude of the intensity stretch. Optical spectroscopic techniques, including Fourier transform infrared (FT-IR) and ultraviolet circular dichroism (UVCD), are useful for studying biological molecules, as they are more widely applicable than the high-resolution method of X-ray crystallography. Databases of protein spectra are currently being developed using these spectroscopic techniques for fold recognition and structural quantification.
Original language | English |
---|---|
Pages (from-to) | 542-544 |
Number of pages | 2 |
Journal | Applied Spectroscopy |
Volume | 59 |
Issue number | 4 |
DOIs | |
Publication status | Published - Apr 2005 |
Keywords
- Data mining
- Intensity stretch
- Normalization
- Scale canceling
- Standards