VAMS vs intraveinous

This repository contains the documents defining and describing the analysis performed in this publication Volani et al. DOI: 10.3390/metabo13020146.

Comparison of capillary whole blood Volumetric Absorbtive Microsampling to intravenous whole blood sampling to intravenous plasma to intravenous red blood cells (RBC)

All data files as well as supporting information files (at the moment only injection sequence) are on the biobank nas under the folder 'data/vams2018'

In the same folder there is also files that contain the injection order in a sub-folder called 'csv'.

The file names follow a naming format convention although the pool samples break the format to a lesser degree. The format is: date-of-analysis_sampleid_matrixid_replicate_polarity. SampleID is just a running number, matrixID is one of a few values: '1', '2', '3', 'RBC' and 'POOLxx'. The xx in the 'POOLxx' indicates the injection order of the pooled sample, while 1, 2, and 3 correspond to 'plasma', 'intravenous whole blood', and 'capillary whole blood' respectively. RBC and POOL should be self explanatory.

Note that most of the files other than the peack picking and feature identification rely on R objects that are generated in the feature identification script. If you haven't run that then chances are the other ones will not work.

News

This is version 3:

• use new adduct information and retention times for standards defined by Mar.
• use integrate = 2 and peakwidth = c(2, 20) for the peak detection.)

Files and structure of the analysis

Data preprocessing and normalization:

• vams_preprocessing.Rmd: Defines the pre-processing of the positive polarity data (chromatographic peak detection, alignment and correspondence).
• vams_preprocessing_neg.Rmd: pre-processing of the negative polarity data.
• vams_normalization.Rmd: performs the normalization of the feature abundances (positive polarity).
• vams_normaliziation_neg.Rmd: normalization of the negative polarity data.

Differences between matrices:

• semi_t_matrices_pos.Rmd: semi-targeted analysis of the matrix specific effects (positive polarity).
• semi_t_matrices_neg.Rmd: semi-targeted analysis of matrix specific effects for negative polarity data.

Differences between male and female individuals:

• semi_t_sex_pos.Rmd: semi-targeted analysis to identify features with significant abundances between male and female participants.
• semi_t_sex_neg.Rmd: semi-targeted analysis to identify features with significant abundances between male and female participants, negative polarity data.
• untargeted_sex_pos.Rmd: untargeted analysis for difference male/female.
• untargeted_sex_neg.Rmd: untargeted analysis for difference male/female, negative polarity.

Required packages and setup

The analysis requires a recent version of R (>= 3.6.0) and the following R packages. All required R packages can be installed with the code below.

install.packages("BiocManager")
BiocManager::install(c("BiocStyle",
                       "xcms",
                       "RColorBrewer",
                       "pander",
                       "doParallel",
                       "magrittr",
                       "pheatmap",
                       "DESeq2",
                       "edgeR",
                       "NormalizeMets",
                       "MetNorm",
                       "ruv",
                       "SummarizedExperiment",
                       "UpSetR",
                       "limma",
                       "writexl",
                       "devtools"))
#' Install from github if versions are "too old"
if (packageVersion("MSnbase") < "2.9.3")
    devtools::install_github("lgatto/MSnbase")
if (packageVersion("xcms") < "3.5.2")
    devtools::install_github("sneumann/xcms")

Raw mzML data location

mzML files with the raw spectrum data are available from the calculation clusters in /data/massspec/mzML/.