Table 2 Determination of unknown samples using two distinct approaches
From: Raman microspectroscopy fingerprinting of organoid differentiation state
(A) Unknown organoid assignments by evaluating multiple peaks: nine peaks individually evaluated for each unknown organoid dataset, the peak assignment determined by mean proximity, with inclusion of significantly overlapping ranges noted in parentheses (see individual peak graphs in Fig. 4). Final unknown assignment determined by simple majority | |||
Peak wavenumber (cm−1) | Unknown 1 | Unknown 2 | Unknown 3 |
569 | Control | Control | EGF (FGF2) |
621 | FGF2 (EGF) | Control | FGF2 |
676 | Control | Control | EGF (FGF2) |
1124 | FGF2 | Control | EGF (FGF2) |
1248 | FGF2 | EGF (control) | FGF2 (EGF) |
1335 | FGF2 | Control (EGF) | EGF |
1446 | FGF2 | Control (EGF) | EGF |
1654 | FGF2 | Control | Control |
1926 | FGF2 | EGF | Control (EGF) |
Final assignment based on multiple peak analysis→ | Unknown 1 = FGF2 (seven out of nine individual peaks assigned as FGF2) | Unknown 2 = control (seven out of nine individual peaks assigned as control) | Unknown 3 = EGF (five out of nine individual peaks assigned as EGF) |
(B) Unknown organoid assignments by SVD analysis: spectra of unknown organoid dataset subjected to SVD analysis with known samples. Line formed between clusters runs perpendicular to the mean of each cluster (Fig. 5). Assignment of unknown based on percent of unknown spectra distribution with known samples (majority of population determines assignment, Fig. 5) | |||
Final assignment based on SVD analysis→ | Unknown 1 = FGF2 →100% of unknown 1-spectrum cluster in FGF2 region compared with EGF (Additional file 1: Fig. S2 shows all comparisons) | Unknown 2 = control →90% of unknown 2-spectrum cluster in control region compared with EGF (Additional file 1: Fig. S3 shows all comparisons) | Unknown 3 = EGF →100% of unknown 3-spectrum cluster in EGF region compared with control (Additional file 1: Fig. S4 shows all comparisons) |
(C) Actual assignment of unknown organoid datasets (unblinding of samples) | |||
|  | Unknown 1 = FGF2 | Unknown 2 = control | Unknown 3 = EGF |