README.md

LOBSTAHS

This is the home of the R package LOBSTAHS, developed in the Van Mooy Lab at Woods Hole Oceanographic Institution. LOBSTAHS (Lipid and Oxylipin Biomarker Screening Through Adduct Hierarchy Sequences) is a multifunction package for screening, annotation, and putative identification of mass spectral features in large, HPLC-MS lipid datasets. LOBSTAHS provides functions for easy creation of custom databases containing entries for a wide range of lipids, oxidized lipids, and oxylipins. Databases are generated using an automated computational approach based on structural criteria specified by the user in simple text tables/spreadsheets. The package's default databases contain roughly 25,000 compounds. Installation instructions and some general information about LOBSTAHS are given below. The current production version of LOBSTAHS is available via Bioconductor.

The primary repository at https://github.com/vanmooylipidomics/LOBSTAHS will always contain the latest development version of LOBSTAHS. Changes are pushed as frequently as possible to Bioconductor, where they will appear in the read-only mirror at https://github.com/Bioconductor-mirror/LOBSTAHS.

A comprehensive vignette contains detailed, step-by-step user instructions and examples of package functions illustrated with a model dataset.

LOBSTAHS and all accompanying written materials are copyright (c) 2015-2021, by the following current and former members of the Van Mooy Laboratory group at Woods Hole Oceanographic Institution: James R. Collins, Bethanie R. Edwards, Daniel Lowenstein, Henry Holm, Helen F. Fredricks, and Benjamin A.S. Van Mooy. LOBSTAHS is provided under the GNU General Public License and subject to terms of reuse as specified therein.

Citations

LOBSTAHS is described in:

Collins, J.R., B.R. Edwards, H.F. Fredricks, and B.A.S. Van Mooy. 2016. LOBSTAHS: An adduct-based lipidomics strategy for discovery and identification of oxidative stress biomarkers. Analytical Chemistry 88:7154-7162; doi:10.1021/acs.analchem.6b01260

As part of Bioconductor, the package also now has its own dedicated DOI: doi:10.18129/B9.bioc.LOBSTAHS

Please use both of these when citing LOBSTAHS in your own work.

Adduct hierarchy and retention time window data for lipid classes BLL, PDPT, vGSL, sGSL, hGSL, hapGSL, and hapCER (package versions ≥ 1.1.2) are as described in:

Hunter J.E., M.J. Frada, H.F. Fredricks, A. Vardi, and B.A.S. Van Mooy. 2015. Targeted and untargeted lipidomics of Emiliania huxleyi viral infection and life cycle phases highlights molecular biomarkers of infection, susceptibility, and ploidy. Frontiers in Marine Science 2:81; doi:10.3389/fmars.2015.00081

and

Fulton, J.M., H.F. Fredricks, K.D. Bidle, A. Vardi, B.J. Kendrick, G.R. DiTullio, and B.A.S. Van Mooy. 2014. Novel molecular determinants of viral susceptibility and resistance in the lipidome of Emiliania huxleyi. Environmental Microbiology 16(4):1137-1149; doi:10.1111/1462-2920.12358

Installation

Install current production version

Users can the current production version of LOBSTAHS from Bioconductor by following the directions here (under "Installation"). The Bioconductor installation function will prompt you to install the latest versions of some other packages on which LOBSTAHS depends.


if (!requireNamespace("BiocManager", quietly=TRUE))
    install.packages("BiocManager")
BiocManager::install()

Note: At this point, you should verify that you have, in fact, installed the most current version of Bioconductor. You may be required to install the latest version of R itself before you can obtain the latest version of Bioc. If you fail to ensure you are running the latest version of Bioconductor at this point, you won't necessarily be able to install the most up-to-date version of LOBSTAHS since the Bioc installer will only provide you will the most recent versions of packages attached to that version of Bioconductor. (If you are confused by any of this, check out this help section. Or, just install the current development version directly from GitHub using the directions, below.) Once you are satisfied that you're running the latest version of Bioconductor, you can then install LOBSTAHS:


if (!requireNamespace("BiocManager", quietly=TRUE))
    install.packages("BiocManager")
BiocManager::install("LOBSTAHS")

Users are also encouraged to download the PtH2O2lipids data package, which can be used for familiarization with the software.

Install "no warranties" development version with latest features

Following these directions, you will install the latest version of the software (including the latest default databases) from the files present in this GitHub repository. Some features may be unstable.

Install dependencies


if (!requireNamespace("BiocManager", quietly=TRUE))
    install.packages("BiocManager")
BiocManager::install("xcms")
BiocManager::install("CAMERA")

Install RTools

For windows: Download and install RTools from http://cran.r-project.org/bin/windows/Rtools/

For Unix: Install the R-development-packages (r-devel or r-base-dev)

Install packages needed for installation from Github:


install.packages("devtools")

Install development version of LOBSTAHS from the GitHub repository


library("devtools")
install_github("vanmooylipidomics/LOBSTAHS", build_vignettes = TRUE) 

The build_vignettes = TRUE argument is required if rendering of the full vignette is desired (recommended). install_github() does not render vignettes by default.

LOBSTAHS dataset preparation in xcms and CAMERA

See this comprehensive vignette for a much more detailed treatment of package functionality.

HPLC-MS data must be assembled into a CAMERA xsAnnotate object prior to analysis with LOBSTAHS. Scripts are provided in the Van Mooy Lab Lipidomics Toolbox for preparing HPLC-ESI-MS data from an Orbitrap Exactive mass spectrometer in xcms and then CAMERA. Or, the user can use his/her own implementation of xcms and CAMERA. We have successfully used the IPO package to optimize xcms and CAMERA settings for a variety of datasets.

In silico database generation

In silico data for a wide range of lipids, oxidized lipids, and oxylipins are generated from user-supplied structural criteria using the database generation function generateLOBdbase(). The function pairs these in silico data with empirically-determined adduct ion abundance rankings for the major lipid classes. Users can generate their own matrices of structural property ranges to be considered during a generateLOBdbase() simulation using the Microsoft Excel spreadsheet templates provided in https://github.com/vanmooylipidomics/LOBSTAHS/tree/master/inst/doc/xlsx. (Users should generate .csv files from this Excel spreadsheets.) Ranges of values can be specified for:

Alternatively, users may load the LOBSTAHS default databases. These contain entries for a wide range of intact polar diacylglycerols (IP-DAG), triacylglycerols (TAG), polyunsaturated aldehydes (PUAs), free fatty acids (FFA), and common photosynthetic pigments. In addition, the latest LOBSTAHS release includes support for lyso lipids under an "IP_MAG" species class and certain glycosphingolipids, ceramides, betaine-like lipids (BLL), bile salts, wax esters, cholesterols, mass spectral contaminants, and quinones. Functionality for other lipid classes is added regularly. The default databases (as of August 30, 2021) include 25,741 and 21,063 unique compounds that can be identifed in positive and negative ionization mode data, respectively.

Identification, screening, and annotation using orthogonal criteria

The function doLOBscreen() is then used to assign putative compound identities from these in silico databases to peakgroups in any high-mass accuracy dataset that has been processed using xcms and CAMERA. doLOBscreen then applies a series of user-selected orthogonal screening criteria based on

to evaluate and assign confidence scores to this list of preliminary assignments. During the screening routine, doLOBscreen() rejects assignments that do not meet the specified criteria, identifies potential isomers and isobars, and assigns a variety of annotation codes to assist the user in evaluating the accuracy of each assignment. Results can be extracted and/or exported to file using the getLOBpeaklist() function.

Example dataset

The package PtH2O2lipids contains a example dataset with which users can familiarize themselves with LOBSTAHS. The dataset contains both a CAMERA "xsAnnotate" object and the LOBSTAHS "LOBSet" generated from it using doLOBscreen(). Processing of the dataset is described in: Collins, J.R., B.R. Edwards, H.F. Fredricks, and B.A.S. Van Mooy. 2016. "LOBSTAHS: An adduct-based lipidomics strategy for discovery and identification of oxidative stress biomarkers." Analytical Chemistry 88:7154-7162; doi:10.1021/acs.analchem.6b01260. Please note that lipid identities were assigned to the PtH2O2lipids dataset using an earlier version of the LOBSTAHS database, which included many fewer compounds than the current version.

Code used to generate figures in Collins et al. LOBSTAHS manuscript

Scripts used to generate the figures and many of the tables in the above referenced manuscript can be found at https://github.com/jamesrco/LipidomicsDataViz/tree/master/LOBSTAHS

LOBSTAHS is maintained by Jamie Collins, Henry Holm, Daniel Lowenstein and other collaborators in the Van Mooy Lab at WHOI.



hholm/LOBSTAHS documentation built on March 17, 2024, 1:53 a.m.