Non-destructive prediction of transcriptomic profiles by Raman microscopy

演者:小林 鉱石 先生

2018年07月11日(水) 午後4時~

Abstract: Raman microscopy is an imaging technique that can report on whole-cell molecular compositions in both comprehensive and non-destructive manners. However, molecular compositions of cells are diverse and compounds such as proteins have severe spectral overlaps, making it nearly intractable to assign specific Raman peaks to specific molecules. Instead of pursuing the spectral decomposition, we show that RNA-seq transcriptomic profiles of Schizosaccharomces pombe and Escherichia coli can be computationally linked and predicted from their whole-cell Raman spectra. By employing machine learning methods, our method finds the intrinsic low-dimensional structure of the transcriptomes, and learns a non-linear linkage between those and Raman spectra. Permutation tests show that the probability of accidentally finding the same transcriptome prediction precision level is extremely low (p-value<0.0001), suggesting that the prediction is real. These results demonstrate that cellular Raman spectra could unravel omics information in a non-destructive manner, opening the possibility of conducting a “live-cell omics.”