Impact of imperfect data on protein secondary structure estimates from Far-UV circular dichroism spectra

Christopher Jones
DOI: https://doi.org/10.1016/j.ab.2024.115483
IF: 2.9
2024-05-01
Analytical Biochemistry
Abstract:Circular dichroism [CD] is widely used to rapidly assess protein structure. Deconvolution of the far-UV CD spectrum is widely used to quantify the secondary structural elements [SSEs]. Multiple algorithms are available for this. Imperfections in the experimental CD spectra arising from spectral noise, instrument miscalibration, spectral offsets and non-linearity will impact on the accuracy and precision of derived secondary structure estimates. Analytical validation for use in regulated environments, such as biopharmaceuticals, requires that the impact of imperfect data on these estimates be understood. Limited information on the impact of poor data were available. A series of noise-free simulated spectral datasets with modified intensity, wavelength, noise and intensity linearity and offsets were created from entries in the Protein Circular Dichroism Data Bank. These datasets were analysed using the BeStSel, on-line resource to estimate secondary structure. Data imperfections caused significant change in SSEs, but the spectral range is also important. This study emphasises the importance of analytical method validation and justifiable estimates of uncertainty when reporting results. The datasets created are made available as a resource to validate other secondary structure estimation programs.
biochemistry & molecular biology,biochemical research methods,chemistry, analytical
What problem does this paper attempt to address?