SpaMTP: Refinning Metabolite Annotations

Refinning m/z Metabolite Annotations with SpaMTP

This tutorial highlights how to use SpaMTP to refine m/z annotated with multiple metabolites. SpaMTP implements this in 4 different ways, with these functions:

1. CalculateAnnotationStatistics
2. RefineLipids
3. Pseudo_msms
4. Compare_msms

We will visit each in detail below.

Author: Andrew Causer

Install and Import R Libraries

First we need to import the required libraries for this analysis.

## Install SpaMTP if not previously installed
if (!require("SpaMTP"))
    devtools::install_github("GenomicsMachineLearning/SpaMTP")

#General Libraries
library(SpaMTP)
library(Cardinal)
library(Seurat)
library(dplyr)

#For plotting + DE plots
library(ggplot2)
library(EnhancedVolcano)
library(viridis)

1) Pathway-Based Refinement: CalculateAnnotationStatistics

This function uses correlations in pathway expression (Metabolite only, Gene only or Multi Omic based pathway information) to refine metabolite annotation. Specifically, for each annotated metabolite, the expression profiles for all known pathways associated with this metabolite are calculated. Based on spatial colocalisation between relative pathways and the given m/z mass, each metabolite is then ranked based on Pearson correlation values and the number of significant pathways associated with that metabolite.

Below we will use a public mouse liver dataset with spotted chemicals standards to demonstrate this:

2) Lipid Nomenchlature Refinement

To come!

3) Pseudo MS/MS-Based Refinement

To come!

4) Refinement with Paired Targeted Metabolic Data

To come!