PMC

All hardware are annotated at sides, and arrows indicate laser pathways.

The single-cell entry in the list, its coordinate, its image and the Raman spectrum were shown on the same screen.

Each data point represents the average value based on 20 single-cells as queries.

The background shows the single-cell images and their positions on the plate under microscope, and at foreground the Raman spectra for 5 single-cells are shown.

(A) Instrument Control, (B) Image Analysis, (C) Raman Profiling, (D) Database Update and (E) Database Search.

In single-cell simulation part (top), the supervised method (based on training dataset) is used for the simulation of single-cell’s image, Raman spectrum and position. In the QSpec system (bottom), the simulation of platform shift has also been implemented.

(A) and (B) showed the Pearson correlations for all peaks measured by automatic (X-axis) and manual (Y-axis) methods for two single-cells. (C) and (D) showed the real Raman profiles measured by automatic (blue) and manual (green) methods for the two single-cells in (A) and (B), respectively.

Single-cells from different strains can be clearly separated by PCA analysis based on their Raman spectra.

It is composed of 4 major parts: (A) cell identity information, (B) culturing experiment information for cells, (C) Raman spectroscopy experiment information for cells, (D) cell image, Raman spectrum and position information.

In (B), the cell sorting is started based on single-cell Raman spectra, and the electromagnetic valve could be turned on and off for cell sorting if (i) the ratio of one peak intensity over another is greater than a threshold as defined or (ii) the difference of one peak intensity minus another is greater than a threshold as defined (as annotated in blue rectangles).