Skip to content

This repository is associated with a biorxiv preprint for the Mag-Net EV enrichment method.

License

Notifications You must be signed in to change notification settings

uw-maccosslab/manuscript-mag-net

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

36 Commits
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

Mag-Net

This repo contains most of the input files and the basic analyses used in the manuscript "Mag-Net: Rapid enrichment of membrane-bound particles enables high coverage quantitative analysis of the plasma proteome", which is currently located on bioRxiv under the DOI 10.1101/2023.06.10.544439. Any files not located here are freely and openly accessible on the Mag-Net page of PanoramaWeb. MS data files can also be found on ProteomeXchange under the unique identifier PXD042947.

Data was exported using the Skyline document grid. Analyses were perfomed using Skyline, R scripts, R Markdown files, and Jupyter Notebooks.

Repository Layout

  • KingFisher Methods: Contains instrument methods for KingFisher System

  • bin: Contains scripts used to generate figures and preprocessing

  • csv: Contains csv files used as input and output for the scripts in the bin folder.

  • elibs: Contains elib files from analysis of the AD/PD data.

$~$

Scripts

Scripts are located in the bin folder.

  • main.r: R script that orchestrates data wrangling and normalization:

    • R environment (version >= 3.6.0 recommended).

    • Required R packages:

      • readr
      • dplyr
      • limma
      • pcaMethods
      • R6
    • Ensure all required packages are installed before running the script. Use install.packages("package_name") to install any missing packages.

    • Usage: navigate to the project directory in your R environment and run the main.r script:

       source("main.r")
      
  • Fig6b.r: R script that is used to generate the following figure panels:

    • Figure 6B
  • Fig6c.r: R script that is used to generate the following figure panels:

    • Figure 6C
  • Fig6d.r: R script that is used to generate the following figure panels:

    • Figure 6D
  • Fig7b_Fig7c.ipynb: Jupyter Notebook including machine learning that is used to generate the following figure panels:

    • Figure 7B
    • Supplementary figure S10
  • Fig2c_Fig4a_Fig4b.r: R script that is used to generate the following figure panels:

    • Figure 3C
    • Figure 4A
    • Figure 4B
    • Figure 7A
  • Figs_2C_3A_3B_5B__SuppFigs_S1_S5_S6.Rmd: R Markdown file that is used to generate the following figure panels:

    • Figure 2C
    • Figure 3A
    • Figure 3B
    • Figure 5B
    • Supplementary Figure S1
    • Supplementary Figure S5
    • Supplementary Figure S6
  • Fig_7C__SuppFigs_S7_S8_S9.Rmd: R Markdown file that is used to generate the following figure panels:

    • Figure 7C
    • Supplementary Figure S7
    • Supplementary Figure S8
    • Supplementary Figure S9

$~$

Files

Input files for scripts are located in the csv folder. Any necessary input files that are too large to be uploaded to GitHub are freely accessible on the Mag-Net page of PanoramaWeb.

main.r input files

Skyline Document: Particles-ADPD-Pilot-EncV2score-QuantReport_2023-03-05_17-26-00.sky.zip

  • EV_ADPD_Protein_Total_Areas.csv: Skyline document grid output of protein abundance from AD/PD experiment.
  • TPAD HC_AD_PDD_PD-noMCI Plasma - Metadata.csv: Skyline document grid output of user-defined metadata for AD/PD experiment.

Fig6b.r input files

Skyline Document: Particles-ADPD-Pilot-EncV2score-QuantReport_2023-03-05_17-26-00.sky.zip

  • EV_ADPD_Peptide_Total_Areas.csv: Skyline document grid output of peptide abundance from AD/PD experiment.
  • TPAD HC_AD_PDD_PD-noMCI Plasma - Metadata.csv: Skyline document grid output of user-defined metadata for AD/PD experiment.

$~$

Fig6c.r input files

Skyline Document: Particles-ADPD-Pilot-EncV2score-QuantReport_2023-03-05_17-26-00.sky.zip

  • EV_ADPD_Peptide_Total_Areas.csv: Skyline document grid output of peptide abundance from AD/PD experiment.
  • TPAD HC_AD_PDD_PD-noMCI Plasma - Metadata.csv: Skyline document grid output of user-defined metadata for AD/PD experiment.

$~$

Fig6d.r input files

Skyline Document: Particles-ADPD-Pilot-EncV2score-QuantReport_2023-03-05_17-26-00.sky.zip

  • EV_ADPD_Peptide_Total_Areas.csv: Skyline document grid output of peptide abundance from AD/PD experiment.
  • EV_ADPD_Protein_Total_Areas.csv: Skyline document grid output of protein abundance from AD/PD experiment.
  • TPAD HC_AD_PDD_PD-noMCI Plasma - Metadata.csv: Skyline document grid output of user-defined metadata for AD/PD experiment.

$~$

Fig7b_Fig7c.ipynb input files

Skyline Document: Particles-ADPD-Pilot-EncV2score-QuantReport_2023-03-05_17-26-00.sky.zip

  • 20230812-particlesADPD_pilot-nobatchProt4.csv: Skyline document grid output of protein abundance from AD/PD experiment.
  • TPAD HC_AD_PDD_PD-noMCI Plasma - Metadata.csv: Skyline document grid output of user-defined metadata for AD/PD experiment.

$~$

Fig2c_Fig4a_Fig4b.r input files

Skyline Document: SAX-EV vs OH PAC-Plasma-EncV2score_2023-06-02_10-52-08.sky.zip

  • Sample 14-Particle Distribution.csv: Size distribution of particles isolated from human plasma by Mag-Net (Figure 3C). Measurements obtained using Nanosight NS300 and analyzed in Nanosight Nanoparticle Tracking Analysis software (Malvern Panalytical Ltd).
  • EV and Plasma Protein Total Areas.csv: Protein areas used to generate the panels depicting the dynamic range of the plasma proteome (Figure 4A and 4B).

$~$

Figs_2C_3A_3B_5B__SuppFigs_S1_S5_S6.Rmd input files

Skyline Document: SAX-EV vs OH PAC-Plasma-EncV2score_2023-06-02_10-52-08.sky.zip

  • Protein_List.csv: Table used to extract and label of proteins of interest highlighted in Enrichment/Depletion plots (Figure 3A and 3B), in the volcano plot annotated with specific extracellular vesicle markers and common plamsa proteins. (Supplementary Figure S1), and AD/PD box plots (Figure 7C, Supplementary Figure S7-S9).

  • group-fc-Beads-SAXN-Particles vs HydN-Plasma-EncV2score-QuantReport_2023-03-05_13-29-41.csv: Skyline document grid output comparing SAX bead Mag-Net enriched fraction to Hydroxyl bead SP3-prepared total plasma samples. Used to generate the volcano plot with protein number annotations (Figure 2C), Enrichment/Depletion plots (Figure 3A and 3B), and in the volcano plot annotated with specific extracellular vesicle markers and common plamsa proteins. (Supplementary Figure S1).

Skyline Document: EV-Matrix Matched Cal Curve_2023-05-24_16-33-15.sky.zip

  • MMCC_EV_Eclipse_Protein_Long_ProtAb_NoNorm.csv: Skyline document grid output of protein abundance without normalization applied.

  • MMCC_EV_Eclipse_figuresofmerit_proteins_NoNorm.csv: Limits of quantitation (LOQ) and Limits of Detection (LOD) generated using variation of python script "calculate_loq.py" in github repo iontrap_vs_orbitrap used in Supplementary Figure S5.

  • MMCC_EV_Eclipse_files_meta.csv: Skyline document grid output of user-defined metadata for MMCC experiment.

  • File located only on PanoramaWeb

    • MMCC_EV_Eclipse_Peptide_Long_TAF_NoNorm.csv: File needed for Matrix-Matched Calibration Curve analysis (Figure 5B). Located in the "Reports" folder under Supplementary Files on the Mag-Net page of PanoramaWeb.

Skyline Document: Particles-FreezeThaw-EncV2score-QuantReport_2023-03-05_14-48-57.sky.zip

  • File located only on PanoramaWeb

    • FT_TICnormalized_LongForm_Peptide_NormArea_Long.csv" Skyline document grid output of peptide normalized area without normalization applied (Supplementary Figure S6). Located in the "Reports" folder under Supplementary Files on the Mag-Net page of PanoramaWeb.
  • FT_LongForm_Reps_Meta.csv: Skyline document grid output of user-defined metadata for freeze-thaw experiment.

$~$

Fig_7C__SuppFigs_S7_S8_S9.Rmd input files

Skyline Document: Particles-ADPD-Pilot-EncV2score-QuantReport_2023-03-05_17-26-00.sky.zip

  • 20230812-particlesADPD_pilot-batchadj-Prot.csv: Normalized protein abundance generated in post-processing.

  • mag-net_biomarker_rocs-final.csv: ROC values generated in post-processing.

  • TPAD HCN_ADD_PDD_PDCN Plasma - Metadata.csv: Skyline document grid output of user-defined metadata for AD/PD experiment.

  • Protein_List.csv: Table used to extract and label of proteins of interest highlighted in Enrichment/Depletion plots (Figure 3A and 3B), in the volcano plot annotated with specific extracellular vesicle markers and common plamsa proteins. (Supplementary Figure S1), and AD/PD box plots (Figure 7C, Supplementary Figure S7-S9).

$~$

Figure Generation

Details of script used broken down for each figure panel.

  • Figure 1:

    • 1A: Mag-Net approach schematic. Generated in an image processor. No script or files on github.
    • 1B: Mag-Net method schematic. Generated in an image processor. No script or files on github.
  • Figure 2:

    • 2A: Peptide detections using different analytical pipelines. No script or files on github.
    • 2B: Protein detections using different analytical pipelines. No script or files on github.
    • 2C: Volcano plot annotated with protein counts. Panel and protein counts generated in "Figs_2C_3A_3B_5B__SuppFigs_S1_S5_S6.Rmd". Protein counts manually added using image processor.
    • 2D: Image of top 10 Jensen Compartments from Enrichr tool. No script or files on github.
  • Figure 3:

    • 3A: Enriched EV markers. Panel generated in "Figs_2C_3A_3B_5B__SuppFigs_S1_S5_S6.Rmd".
    • 3B: Depleted common plasma proteins. Panel generated in "Figs_2C_3A_3B_5B__SuppFigs_S1_S5_S6.Rmd".
    • 3C: Particle count distribution. Panel generated in "Fig2c_Fig4a_Fig4b.r".
    • 3D: TEM image collected using a JEOL 1230 Transmission Electron Microscope (Peabody, MA). No script or files on github.
  • Figure 4:

    • 4A: Dynamic range of the plasma proteome - plasma digestion. Panel generated in "Fig2c_Fig4a_Fig4b.r" and location of protein text labels manually optimized in image processor for better visualization.
    • 4B: Dynamic range of the plasma proteome - Mag-Net enrichment. Panel generated in "Fig2c_Fig4a_Fig4b.r" and location of protein text labels manually optimized in image processor for better visualization.
  • Figure 5:

    • 5A: MMCC visual summary. Generated in an image processor. No script or files on github.
    • 5B: Half-violin/ box plot of all peptides in MMCC experiment. Panel generated in "Figs_2C_3A_3B_5B__SuppFigs_S1_S5_S6.Rmd".
  • Figure 6:

    • 6A: Schematic describing dementia cohort. Generated in an image processor. No script or files on github.
    • 6B: Plotting peptide detections and FDR levels in each experimental and quality control sample. Panel generated in "Fig6b.r".
    • 6C: Log2 peptide abundance of replicate Mag-Net preparation. Panel generated in "Fig6c.r".
    • 6D: Effect of data processing on CVs. Panel generated in "Fig6d.r".
  • Figure 7:

    • 7A: Heatmap of proteins with ROC > 0.7 from 6 pairwise analyses. Panel generated in "Fig2c_Fig4a_Fig4b.r".
    • 7B: Example receiver operator characteristic (ROC) curve. Panel generated in "Fig7b_Fig7c.ipynb".
    • 7C: Panel of protein-level box plots annotated with ROC and q-value of 10 proteins that were found to be specifically increased in ADD. Panel generated in "Fig_7C__SuppFigs_S7_S8_S9.Rmd".
  • Supplementary Figure S1:

    • Volcano plot annotated with specific extracellular vesicle markers and common plamsa proteins. Panel generated in "Figs_2C_3A_3B_5B__SuppFigs_S1_S5_S6.Rmd".
  • Supplementary Figure S2:

    • Image of volcano plot exported from Skyline comparing hydroxyl and SAX beads on the same unfractionated plasma.
  • Supplementary Figure S3:

    • S3A: Image of volcano plot exported from Skyline comparing total plasma digested using hydroxyl beads and the same plasma processed with Mag-Net using hydroxyl beads. No script or files on github.
    • S3B: Image of top 10 Jensen Compartments. No script or files on github.
  • Supplementary Figure S4:

    • S4A: Image of top 10 GO terms in Hydroxyl bead Mag-Net capture. No script or files on github.
    • S4B: Image of volcano plot exported from Skyline comparing plasma processed with Mag-Net using both hydroxyl and SAX beads. No script or files on github.
    • S4C: Image of top 10 GO terms in SAX bead Mag-Net capture. No script or files on github.
  • Supplementary Figure S5:

    • S5A: MMCC visual summary. Generated in an image processor. No script or files on github.
    • S5B: CD9 - Protein abundance of replicate injections exported from Skyline document grid. Limit of quantitation (LOQ), and Limit of Detection (LOD) calculated using variation of python script "calculate_loq.py" in github repo iontrap_vs_orbitrap. Panel generated in "Figs_2C_3A_3B_5B__SuppFigs_S1_S5_S6.Rmd".
    • S5C: NCAM1 - Protein abundance of replicate injections exported from Skyline document grid. Limit of quantitation (LOQ), and Limit of Detection (LOD) calculated using variation of python script "calculate_loq.py" in github repo iontrap_vs_orbitrap. Panel generated in "Figs_2C_3A_3B_5B__SuppFigs_S1_S5_S6.Rmd".
    • S5D: CD9 - Protein abundance of replicate injections exported from Skyline document grid. Limit of quantitation (LOQ), and Limit of Detection (LOD) calculated using variation of python script "calculate_loq.py" in github repo iontrap_vs_orbitrap. Panel generated in "Figs_2C_3A_3B_5B__SuppFigs_S1_S5_S6.Rmd".
  • Supplementary Figure S6:

    • Half-violin/ box plot of all peptides in FT experiment. Panel generated in "Figs_2C_3A_3B_5B__SuppFigs_S1_S5_S6.Rmd".
  • Supplementary Figure S7:

    • Panel of protein-level box plots annotated with ROC and q-value of 12 proteins that were found to be specifically increased in ADD and involved in vesicle mediated transport. Panel generated in "Fig_7C__SuppFigs_S7_S8_S9.Rmd".
  • Supplementary Figure S8:

    • Panel of protein-level box plots annotated with ROC and q-value of 8 proteins that were found to be specifically increased in ADD and are involved in lipid metabolism. Panel generated in "Fig_7C__SuppFigs_S7_S8_S9.Rmd".
  • Supplementary Figure S9:

    • Panel of protein-level box plots annotated with ROC and q-value of 16 proteins that were found to be specifically increased in ADD and are involved in Ubiquitin-proteasome mediated protein degradation. Panel generated in "Fig_7C__SuppFigs_S7_S8_S9.Rmd".
  • Supplementary Figure S10:

    • Panel of 5 ROC curves illustrating development of an SVM classifier. Panel generated in "Fig7b_Fig7c.ipynb".

About

This repository is associated with a biorxiv preprint for the Mag-Net EV enrichment method.

Topics

Resources

License

Stars

Watchers

Forks