LAMP - Liverpool Annotation of metabolites using Mass sPectrometry

Introduction

Untargeted metabolomics studies routinely apply liquid chromatography-mass spectrometry to acquire data for hundreds or low thousands of metabolites and exposome-related (bio)chemicals. The annotation or higher-confidence identification of metabolites and biochemicals can apply multiple different data types (1) chromatographic retention time, (2) the mass-to-charge (m/z) ratio of ions formed during electrospray ionisation for the structurally intact metabolite or (bio)chemical and (3) fragmentation mass spectra derived from MS/MS or MS^n^ experiments.

Commonly, the mass-to-charge (m/z) ratio of ions formed during electrospray ionisation for the structurally intact metabolite are applied as a first step in the annotation process. Importantly, a single metabolite can be detected as multiple different ion types (adducts, isotopes, in-source fragments, oligomers) and grouping together of features representing the same metabolite or biochemical can decrease the number of false positive annotations. The Liverpool Annotation of metabolites using Mass sPectrometry (LAMP) is a Python package and an easy-to-use software for feature grouping and metabolite annotation using MS1 data only. LAMP groups features based on chromatographic retention time similarity and positive response-based correlations across multiple biological samples. Genome-scale metabolic models are the source of metabolites applied in the standard reference files though any source of metabolites can be used (e.g. HMDB or LIPIDMAPS). The m/z differences related to in-source fragments, adducts, isotopes, oligomers and charge states can be user-defined in the reference file.

Installation

PyPI

To install from PyPI via pip, use the distribution name lamps:

pip install lamps

This is the preferred installation method.

Conda

LAMP is in Bioconda channel and use the following to install for conda:

conda install -c bioconda lamps

Source

Install directly from GitHub:

pip install git+https://github.com/wanchanglin/lamp.git

Update directly from GitHub branch dev:

pip install git+https://github.com/wanchanglin/lamp.git@dev --upgrade --no-deps --force-reinstall

Usages

For end users, LAMP provides command line and graphical user interfaces.

$ lamp --help
Executing lamp version 1.0.4.
usage: lamp [-h] {cli,gui} ...

Compounds Annotation of LC-MS data

positional arguments:
  {cli,gui}
    cli       Annotate metabolites in CLI.
    gui       Annotate metabolites in GUI.

options:
  -h, --help  show this help message and exit

Command line interface (CLI)

Use the follow command line to launch CLI mode: :

$ lamp cli <arg_lists>

Here is an example: :

lamp cmd \
  --sep "tab" \
  --input-data "./data/df_pos_3.tsv" \
  --col-idx "1, 2, 3, 4" \
  --add-path "" \
  --ref-path "" \
  --ion-mode "pos" \
  --cal-mass \
  --thres-rt "1.0" \
  --thres-corr "0.5" \
  --thres-pval "0.05" \
  --method "pearson" \
  --positive \
  --ppm "5.0" \
  --save-db \
  --save-mr \
  --db-out "./res/test.db" \
  --sr-out "./res/test_s.tsv" \
  --mr-out "./res/test_m.tsv"

Graphical user interface (GUI)

$ lamp gui

Authors

Wanchang Lin (Wanchang.Lin@liverpool.ac.uk), The University of Liverpool
Warwick Dunn (Warwick.Dunn@liverpool.ac.uk), The University of Liverpool