Supplementary Data from Picosecond Infrared Laser Desorption Mass Spectrometry Identifies Medulloblastoma Subgroups on Intrasurgical Timescales

Authors:
  • MWMichael Woolman
    Michael Woolman
  • CKClaudia M. Kuzan-Fischer
    Claudia M. Kuzan-Fischer
  • IFIsabelle Ferry
    Isabelle Ferry
  • TKTaira Kiyota
    Taira Kiyota
  • BLBetty Luu
    Betty Luu
  • MWMegan Wu
    Megan Wu
  • DMDavid G. Munoz
    David G. Munoz
  • SDSunit Das
    Sunit Das
  • AAAhmed Aman
    Ahmed Aman
  • MTMichael D. Taylor
    Michael D. Taylor
  • JRJames T. Rutka
    James T. Rutka
  • HGHoward J. Ginsberg
    Howard J. Ginsberg

Abstract
Table S1. Sampling events, signal strength and signal persistence of data points used to build the MB subgroup model; Table S2. MB subgroup classifying peak list from PCA-LDA modeling;Table S3. PCA-LDA prediction results of the blind MB subgrouping test; Figure S1. The collection geometry of PIRL-MS setup used in this study; Figure S2. Representative PIRL-MS spectra of MB subgroups; Figure S3. The PCA-LDA model of the 4 MB subgroups upon application of non-subjective data exclusion criteria; Figure S4. The 'low complexity' PCA-LDA model of the 4 MB subgroups utilizing only the top 44 MB subgroup distinguishing m/z values; Figure S5. Hematoxylin and Eosin (H&E) staining of the misclassified and unclassifiable data points; Figure S6. PCA-LDA analysis of the misclassified/unclassifiable data points; Figure S7. Liquid Chromatography (LC) elution profiles of laser plume and tissue extract; Figure S8. The envisioned use scenarios for PIRL-MS technology. We have also per request of the reviewers included information on identification of MB classifying molecules with LC-MS, Intrasample variability in correct subgroup assignment as well as Sample collection and handling prior to access for PIRL-MS sampling