Validation of biosignatures confirms the informative nature of fossil organic Raman spectra

Abstract

Raman spectroscopy has facilitated rapid progress in the understanding of patterns and processes associated with biomolecule fossilization and revealed the preservation of biological and geological signatures in fossil organic matter. Nonetheless six large-scale statistical studies of Raman spectra of carbonaceous fossils, selected from a number of independent assessments producing similar trends, have been disputed. Alleon et al. (21) applied a wavelet transform analysis in an unconventional way to identify frequency components contributing to two baselined spectra selected from these studies and claimed similarities with a downloaded edge filter transmission spectrum. On the basis of indirect comparisons and qualitative observations they argued that all spectral features detected, including significant mineral peaks, can be equated to edge filter ripples and are therefore artefactual. Alleon et al. (21) extrapolated this conclusion to dispute not only the validity of n>200 spectra in the studies in question, but also the utility of Raman spectroscopy, a well established method, for analysing organic materials in general. Here we test the claims by Alleon et al. (21) using direct spectral comparisons and statistical analyses. We present multiple independent lines of evidence that demonstrate the original, biologically and geologically informative nature of the Raman spectra in question. We demonstrate that the methodological approach introduced by Alleon et al. (21) is unsuitable for assessing the quality of spectra and identifying noise within them. Statistical analyses of large Raman spectral data sets provide a powerful tool in the search for compositional patterns in biomaterials and yield invaluable insights into the history of life.