In MoseleyBioinformaticsLab/ScanCentricPeakCharacterization: Functionality for Characterizing Peaks in Mass Spectrometry in a Scan-Centric Manner

knitr::opts_chunk$set(
  collapse = TRUE,
  comment = "#>"
)

library(ScanCentricPeakCharacterization)

Introduction

So you want to do some peak characterization, using ScanCentricPeakCharacterization, hereafter shortened to SCPC. If you haven't already, you might want to read the publication describing the motivation behind it [@flightScanCentricFrequencyBasedMethod2022].

First things to know:

• Is your data in one of the open formats that mzR and MSnbase support (mzML, mzXML, mzData)?
• Do you have multiple, non-chromatographic scans or acquisitions of data where the same M/Z range is covered?
• Is your data acquired in profile mode?
• Is your data from an ICR or orbitrap type of instrument (Fusion, Fusion Lumos, SolariX, etc)?

If the answers to the above are all true, then you should be OK going forward. If any of them are not true, or your aren't sure, then feel free to contact the package author, Robert Flight , or file an issue on the GitHub repo of this package.

Things to Know

R6 Based

This package is organized around R6 objects, as we have large data that we don't want to worry about making copies of.

Multiprocessing

SCPC is an intensive process. It first detects peaks in every scan, figures out the matched peaks across scans, and then does the full characterization of each peak across the scans. So ideally, you want to enable parallel processing when using this package, if available. In this package, this is enabled via furrr and future_map.

#| label: enable-multicore
#| eval: false
# if in RStudio
# options(parallelly.fork.enable = TRUE)
library(furrr)
plan(multicore)
# this tells SCPC use `furrr::future_map` instead of `purrr::map`
set_internal_map(furrr::future_map)

Logging

SCPC also enables logging to help keep you apprised of what is happening as far as progress and memory usage. The latter is very useful when you have very large sets of data, and you want to make sure that it fits in available memory, or you need to use fewer processes to avoid using all of the RAM on your machine (see Multiprocessing).

It does require that the logger package is installed. You can turn on logging using enable_logging.

#| label: enable-logging
#| eval: false
enable_logging()

Setup

OK, let's assume you've got your data, and you are ready to go. For the following examples, we are going to work with a lipidomics sample acquired on a Thermo-Fisher Tribrid-Fusion,

References

MoseleyBioinformaticsLab/ScanCentricPeakCharacterization documentation built on March 29, 2024, 11:32 p.m.