In IFB-ElixirFr/ProMetIS: Multi-omics phenotyping of the LAT and MX2 knockout mice
knitr::opts_chunk$set(fig.width = 8,
fig.height = 8,
fig.path = 'figures/temp/1_post_processing/',
warning = FALSE)
This vignette describes the post-processing of proteomics datasets (mainly re-formating) and of the metabolomics datasets (normalization including filtering based on quality metrics and signal drift correction; see the 'Normalization of the dataMatrix' section for details; log2 transformation). The input datasets are from the '1_processed' repository while the post-processed datasets are exported into the '2_post_processed' repository.
Note: The MutiDataSet/ExpressionSet Bioconductor framework is used throughout this vignette.
Mice IDs
processed.mset <- phenomis::reading(ProMetIS::processed_dir.c(),
output.c = "set",
report.c = "none")
preclin.eset <- processed.mset[["preclinical"]]
mice_id.df <- Biobase::pData(preclin.eset)
mice_id.df[, "id"] <- NULL
Proteomics
Set names and input files
proteo_files.vc <- vapply(ProMetIS::proteo_sets.vc(),
function(set.c) {
files.vc <- list.files(file.path(ProMetIS::processed_dir.c(), set.c),
pattern = ".xlsx", full.names = TRUE)
files.vc[!grepl("(ProMetIS|prometis)", basename(files.vc))]
}, FUN.VALUE = character(1))
proteo.mset <- MultiDataSet::createMultiDataSet()
MultiDataSet containing the data matrices
for (set.c in ProMetIS::proteo_sets.vc()) {
# dataMatrix
data.df <- as.data.frame(readxl::read_excel(proteo_files.vc[set.c],
sheet = 1),
stringsAsFactors = FALSE)
rownames(data.df) <- data.df[, 1]
data.df[, 1] <- NULL
data.mn <- as.matrix(data.df)
rm(data.df)
mode(data.mn) <- "numeric"
eset <- Biobase::ExpressionSet(assayData = data.mn,
experimentData = Biobase::MIAME(title = set.c))
stopifnot(methods::validObject(eset))
proteo.mset <- MultiDataSet::add_eset(proteo.mset,
eset,
dataset.type = set.c,
GRanges = NA,
overwrite = TRUE,
warnings = FALSE)
}
sampleMetadata
for (set.c in ProMetIS::proteo_sets.vc()) {
eset <- proteo.mset[[set.c]]
proteo_pda.df <- as.data.frame(readxl::read_excel(proteo_files.vc[set.c],
sheet = 2),
stringsAsFactors = FALSE)
rownames(proteo_pda.df) <- gsub(".", "_",
proteo_pda.df[, 1], fixed = TRUE)
sample_names.vc <- Biobase::sampleNames(eset)
if (set.c == "proteomics_liver") {
stopifnot(identical(sort(rownames(proteo_pda.df)),
sort(sample_names.vc)))
# proteo_pda.df <- proteo_pda.df[sample_names.vc, ]
}
proteo_pda.df <- proteo_pda.df[sample_names.vc, ]
# stopifnot(identical(rownames(proteo_pda.df),
# sample_names.vc))
Biobase::pData(eset) <- proteo_pda.df
stopifnot(methods::validObject(eset))
proteo.mset <- MultiDataSet::add_eset(proteo.mset,
eset,
dataset.type = set.c,
GRanges = NA,
overwrite = TRUE,
warnings = FALSE)
}
variableMetadata
for (set.c in ProMetIS::proteo_sets.vc()) {
eset <- proteo.mset[[set.c]]
proteo_fda.df <- as.data.frame(readxl::read_excel(proteo_files.vc[set.c],
sheet = 3),
stringsAsFactors = FALSE)
rownames(proteo_fda.df) <- proteo_fda.df[, 1]
stopifnot(identical(rownames(proteo_fda.df),
Biobase::featureNames(eset)))
proteo_fda.df[, "uniprot_id"] <- sapply(proteo_fda.df[, "accession"],
function(access.c)
unlist(strsplit(access.c,
split = "|",
fixed = TRUE))[2],
USE.NAMES = FALSE)
Biobase::fData(eset) <- proteo_fda.df
stopifnot(methods::validObject(eset))
proteo.mset <- MultiDataSet::add_eset(proteo.mset,
eset,
dataset.type = set.c,
GRanges = NA,
overwrite = TRUE,
warnings = FALSE)
}
Renaming features and re-ordering samples
for (set.c in ProMetIS::proteo_sets.vc()) {
eset <- proteo.mset[[set.c]]
# changing variable IDs to names
feat_names.vc <- paste0(sapply(Biobase::fData(eset)[, "accession"],
function(access.c)
unlist(strsplit(access.c, split = "|",
fixed = TRUE))[2]),
"_",
sapply(Biobase::fData(eset)[, "description"],
function(desc.c) {
if (!is.na(desc.c) && desc.c != "") {
desc.c <- unlist(strsplit(desc.c, split = " OS="))[1]
if (nchar(desc.c) > 25)
desc.c <- paste0(substr(desc.c, 1, 24), ".")
return(desc.c)
}
}))
stopifnot(!any(duplicated(feat_names.vc)))
Biobase::featureNames(eset) <- feat_names.vc
# discarding pools
if (set.c == "proteomics_liver")
eset <- eset[, !grepl("(p|P)ool", Biobase::pData(eset)[, "sample name"])]
# re-ordering samples
if (set.c == "proteomics_liver") {
eset <- eset[, order(as.integer(Biobase::pData(eset)[, "sample name"]))]
} else {
Biobase::sampleNames(eset) <- sapply(Biobase::sampleNames(eset),
function(samp.c) unlist(strsplit(samp.c, split = "_"))[2], USE.NAMES = FALSE)
eset <- eset[, order(as.integer(Biobase::sampleNames(eset)))]
}
# renaming samples
if (set.c == "proteomics_liver") {
Biobase::sampleNames(eset) <- paste0(sapply(Biobase::pData(eset)[, "Condition"],
function(cond.c)
switch(cond.c,
mx2 = "X",
wt = "W",
lat = "L")),
Biobase::pData(eset)[, "sample name"])
} else {
Biobase::sampleNames(eset) <- paste0(sapply(Biobase::pData(eset)[, "Condition"],
function(cond.c)
switch(cond.c,
Mx2 = "X",
CONT = "W",
LAT = "L")),
Biobase::sampleNames(eset))
}
mice_konum.vc <- substr(rownames(mice_id.df), 1, 4)
mice_koid.vc <- rownames(mice_id.df)
if (set.c == "proteomics_plasma") {
stopifnot(all(Biobase::sampleNames(eset) %in% mice_konum.vc))
mice_koid.vc <- mice_koid.vc[mice_konum.vc %in% Biobase::sampleNames(eset)]
mice_konum.vc <- mice_konum.vc[mice_konum.vc %in% Biobase::sampleNames(eset)]
}
stopifnot(identical(Biobase::sampleNames(eset),
mice_konum.vc))
Biobase::sampleNames(eset) <- mice_koid.vc
stopifnot(methods::validObject(eset))
proteo.mset <- MultiDataSet::add_eset(proteo.mset,
eset,
dataset.type = set.c,
GRanges = NA,
overwrite = TRUE,
warnings = FALSE)
}
Adding meta data (gene, mouse_id, sex)
for (set.c in ProMetIS::proteo_sets.vc()) {
eset <- proteo.mset[[set.c]]
Biobase::pData(eset) <- cbind.data.frame(mice_id.df[Biobase::sampleNames(eset), ],
Biobase::pData(eset)[, setdiff(Biobase::varLabels(eset),
colnames(mice_id.df))],
stringsAsFactors = FALSE)
stopifnot(methods::validObject(eset))
proteo.mset <- MultiDataSet::add_eset(proteo.mset,
eset,
dataset.type = set.c,
GRanges = NA,
overwrite = TRUE,
warnings = FALSE)
}
Variable metadata: including new mice ids in the column names (count, abundance, etc.)
for (set.c in ProMetIS::proteo_sets.vc()) {
eset <- proteo.mset[[set.c]]
samp.vc <- Biobase::sampleNames(eset)
pdata.df <- Biobase::pData(eset)
fvar.vc <- Biobase::fvarLabels(eset)
for (i in seq_along(samp.vc)) {
samp.c <- samp.vc[i]
samp_id.c <- gsub("abundance_", "", pdata.df[i, "id"])
fvar_id.vi <- grep(paste0(".+", samp_id.c), fvar.vc)
fvar.vc[fvar_id.vi] <- gsub(switch(set.c,
proteomics_liver = "mgf",
proteomics_plasma = samp_id.c),
samp.c, fvar.vc[fvar_id.vi])
}
Biobase::fvarLabels(eset) <- fvar.vc
stopifnot(methods::validObject(eset))
proteo.mset <- MultiDataSet::add_eset(proteo.mset,
eset,
dataset.type = set.c,
GRanges = NA,
overwrite = TRUE,
warnings = FALSE)
}
Re-ordering metadata ('supp_' columns at the end)
proteo.mset <- ProMetIS:::metadata_select(proteo.mset, step.c = "2_post_processed")
Saving
for (set.c in ProMetIS::proteo_sets.vc()) {
eset <- proteo.mset[[set.c]]
# an 'id' extra column has been automatically created in the sampleMetadata by
# 'MultiDataSet' and must be removed before saving
pdata.df <- Biobase::pData(eset)
pdata.df[, "id"] <- NULL
Biobase::pData(eset) <- pdata.df
phenomis::writing(eset,
file.path(gsub(ProMetIS::data_dir.c(),
"../../ProMetIS/inst/extdata",
ProMetIS::post_processed_dir.c()),
set.c),
overwrite.l = TRUE)
}
Metabolomics
Loading
metabo.mset <- phenomis::reading(ProMetIS::processed_dir.c(),
subsets.vc = ProMetIS::metabo_sets.vc(),
output.c = "set",
report.c = "none")
Normalization of the dataMatrix
The workflow consists of the following steps:
| step | liver_c18hypersil_pos | liver_hilic_neg | plasma_c18hypersil_pos | plasma_hilic_neg | plasma_c18acquity_pos | plasma_c18acquity_neg |
|-----------|-----------|-----------|-----------|-----------|-----------|-----------|
| 0.4 \<= RT \<= 22 | | | | | X | X |
| blank/samp \<= 0.33 | X | X | X | X | X | X |
| poolDil: cor >= 0.7 | X | X | X | X | | |
| drift loess, span=1 | none | none | pool | pool | sample | sample |
| NA \<= 20% & var > 0 | X | X | X | X | X | X |
| pool CV \<= 0.3 | X | X | X | X | X | X |
| poolCV/sampCV \<= 1 | X | X | X | X | X | X |
| corr, dRT, dmz | X | X | X | X | X | X |
metabo.mset <- ProMetIS:::.metabo_postprocessing(metabo.mset = metabo.mset,
drift_correct.c = "prometis")
Formatting and ordering sample names
metabo.mset <- ProMetIS:::.format_metabonames(metabo.mset = metabo.mset,
mice_id.df = mice_id.df)
Log2 transformation
metabo.mset <- phenomis::transforming(metabo.mset, report.c = "none")
Re-ordering metadata ('supp_' columns at the end)
metabo.mset <- ProMetIS:::metadata_select(metabo.mset,
step.c = "2_post_processed")
Saving
for (set.c in ProMetIS::metabo_sets.vc()) {
eset <- metabo.mset[[set.c]]
# an 'id' extra column has been automatically created in the sampleMetadata by
# 'MultiDataSet' and must be removed before saving
pdata.df <- Biobase::pData(eset)
pdata.df[, "id"] <- NULL
Biobase::pData(eset) <- pdata.df
phenomis::writing(eset,
file.path(gsub(ProMetIS::data_dir.c(),
"../../ProMetIS/inst/extdata",
ProMetIS::post_processed_dir.c()),
set.c),
overwrite.l = TRUE)
}
IFB-ElixirFr/ProMetIS documentation built on May 21, 2024, 8:02 p.m.
R Package Documentation
Browse R Packages
We want your feedback!
Note that we can't provide technical support on individual packages. You should contact the package authors for that.
knitr::opts_chunk$set(fig.width = 8, fig.height = 8, fig.path = 'figures/temp/1_post_processing/', warning = FALSE)
This vignette describes the post-processing of proteomics datasets (mainly re-formating) and of the metabolomics datasets (normalization including filtering based on quality metrics and signal drift correction; see the 'Normalization of the dataMatrix' section for details; log2 transformation). The input datasets are from the '1_processed' repository while the post-processed datasets are exported into the '2_post_processed' repository.
Note: The MutiDataSet/ExpressionSet Bioconductor framework is used throughout this vignette.
Mice IDs
processed.mset <- phenomis::reading(ProMetIS::processed_dir.c(), output.c = "set", report.c = "none") preclin.eset <- processed.mset[["preclinical"]] mice_id.df <- Biobase::pData(preclin.eset) mice_id.df[, "id"] <- NULL
Proteomics
Set names and input files
proteo_files.vc <- vapply(ProMetIS::proteo_sets.vc(), function(set.c) { files.vc <- list.files(file.path(ProMetIS::processed_dir.c(), set.c), pattern = ".xlsx", full.names = TRUE) files.vc[!grepl("(ProMetIS|prometis)", basename(files.vc))] }, FUN.VALUE = character(1)) proteo.mset <- MultiDataSet::createMultiDataSet()
MultiDataSet containing the data matrices
for (set.c in ProMetIS::proteo_sets.vc()) { # dataMatrix data.df <- as.data.frame(readxl::read_excel(proteo_files.vc[set.c], sheet = 1), stringsAsFactors = FALSE) rownames(data.df) <- data.df[, 1] data.df[, 1] <- NULL data.mn <- as.matrix(data.df) rm(data.df) mode(data.mn) <- "numeric" eset <- Biobase::ExpressionSet(assayData = data.mn, experimentData = Biobase::MIAME(title = set.c)) stopifnot(methods::validObject(eset)) proteo.mset <- MultiDataSet::add_eset(proteo.mset, eset, dataset.type = set.c, GRanges = NA, overwrite = TRUE, warnings = FALSE) }
sampleMetadata
for (set.c in ProMetIS::proteo_sets.vc()) { eset <- proteo.mset[[set.c]] proteo_pda.df <- as.data.frame(readxl::read_excel(proteo_files.vc[set.c], sheet = 2), stringsAsFactors = FALSE) rownames(proteo_pda.df) <- gsub(".", "_", proteo_pda.df[, 1], fixed = TRUE) sample_names.vc <- Biobase::sampleNames(eset) if (set.c == "proteomics_liver") { stopifnot(identical(sort(rownames(proteo_pda.df)), sort(sample_names.vc))) # proteo_pda.df <- proteo_pda.df[sample_names.vc, ] } proteo_pda.df <- proteo_pda.df[sample_names.vc, ] # stopifnot(identical(rownames(proteo_pda.df), # sample_names.vc)) Biobase::pData(eset) <- proteo_pda.df stopifnot(methods::validObject(eset)) proteo.mset <- MultiDataSet::add_eset(proteo.mset, eset, dataset.type = set.c, GRanges = NA, overwrite = TRUE, warnings = FALSE) }
variableMetadata
for (set.c in ProMetIS::proteo_sets.vc()) { eset <- proteo.mset[[set.c]] proteo_fda.df <- as.data.frame(readxl::read_excel(proteo_files.vc[set.c], sheet = 3), stringsAsFactors = FALSE) rownames(proteo_fda.df) <- proteo_fda.df[, 1] stopifnot(identical(rownames(proteo_fda.df), Biobase::featureNames(eset))) proteo_fda.df[, "uniprot_id"] <- sapply(proteo_fda.df[, "accession"], function(access.c) unlist(strsplit(access.c, split = "|", fixed = TRUE))[2], USE.NAMES = FALSE) Biobase::fData(eset) <- proteo_fda.df stopifnot(methods::validObject(eset)) proteo.mset <- MultiDataSet::add_eset(proteo.mset, eset, dataset.type = set.c, GRanges = NA, overwrite = TRUE, warnings = FALSE) }
Renaming features and re-ordering samples
for (set.c in ProMetIS::proteo_sets.vc()) { eset <- proteo.mset[[set.c]] # changing variable IDs to names feat_names.vc <- paste0(sapply(Biobase::fData(eset)[, "accession"], function(access.c) unlist(strsplit(access.c, split = "|", fixed = TRUE))[2]), "_", sapply(Biobase::fData(eset)[, "description"], function(desc.c) { if (!is.na(desc.c) && desc.c != "") { desc.c <- unlist(strsplit(desc.c, split = " OS="))[1] if (nchar(desc.c) > 25) desc.c <- paste0(substr(desc.c, 1, 24), ".") return(desc.c) } })) stopifnot(!any(duplicated(feat_names.vc))) Biobase::featureNames(eset) <- feat_names.vc # discarding pools if (set.c == "proteomics_liver") eset <- eset[, !grepl("(p|P)ool", Biobase::pData(eset)[, "sample name"])] # re-ordering samples if (set.c == "proteomics_liver") { eset <- eset[, order(as.integer(Biobase::pData(eset)[, "sample name"]))] } else { Biobase::sampleNames(eset) <- sapply(Biobase::sampleNames(eset), function(samp.c) unlist(strsplit(samp.c, split = "_"))[2], USE.NAMES = FALSE) eset <- eset[, order(as.integer(Biobase::sampleNames(eset)))] } # renaming samples if (set.c == "proteomics_liver") { Biobase::sampleNames(eset) <- paste0(sapply(Biobase::pData(eset)[, "Condition"], function(cond.c) switch(cond.c, mx2 = "X", wt = "W", lat = "L")), Biobase::pData(eset)[, "sample name"]) } else { Biobase::sampleNames(eset) <- paste0(sapply(Biobase::pData(eset)[, "Condition"], function(cond.c) switch(cond.c, Mx2 = "X", CONT = "W", LAT = "L")), Biobase::sampleNames(eset)) } mice_konum.vc <- substr(rownames(mice_id.df), 1, 4) mice_koid.vc <- rownames(mice_id.df) if (set.c == "proteomics_plasma") { stopifnot(all(Biobase::sampleNames(eset) %in% mice_konum.vc)) mice_koid.vc <- mice_koid.vc[mice_konum.vc %in% Biobase::sampleNames(eset)] mice_konum.vc <- mice_konum.vc[mice_konum.vc %in% Biobase::sampleNames(eset)] } stopifnot(identical(Biobase::sampleNames(eset), mice_konum.vc)) Biobase::sampleNames(eset) <- mice_koid.vc stopifnot(methods::validObject(eset)) proteo.mset <- MultiDataSet::add_eset(proteo.mset, eset, dataset.type = set.c, GRanges = NA, overwrite = TRUE, warnings = FALSE) }
Adding meta data (gene, mouse_id, sex)
for (set.c in ProMetIS::proteo_sets.vc()) { eset <- proteo.mset[[set.c]] Biobase::pData(eset) <- cbind.data.frame(mice_id.df[Biobase::sampleNames(eset), ], Biobase::pData(eset)[, setdiff(Biobase::varLabels(eset), colnames(mice_id.df))], stringsAsFactors = FALSE) stopifnot(methods::validObject(eset)) proteo.mset <- MultiDataSet::add_eset(proteo.mset, eset, dataset.type = set.c, GRanges = NA, overwrite = TRUE, warnings = FALSE) }
Variable metadata: including new mice ids in the column names (count, abundance, etc.)
for (set.c in ProMetIS::proteo_sets.vc()) { eset <- proteo.mset[[set.c]] samp.vc <- Biobase::sampleNames(eset) pdata.df <- Biobase::pData(eset) fvar.vc <- Biobase::fvarLabels(eset) for (i in seq_along(samp.vc)) { samp.c <- samp.vc[i] samp_id.c <- gsub("abundance_", "", pdata.df[i, "id"]) fvar_id.vi <- grep(paste0(".+", samp_id.c), fvar.vc) fvar.vc[fvar_id.vi] <- gsub(switch(set.c, proteomics_liver = "mgf", proteomics_plasma = samp_id.c), samp.c, fvar.vc[fvar_id.vi]) } Biobase::fvarLabels(eset) <- fvar.vc stopifnot(methods::validObject(eset)) proteo.mset <- MultiDataSet::add_eset(proteo.mset, eset, dataset.type = set.c, GRanges = NA, overwrite = TRUE, warnings = FALSE) }
Re-ordering metadata ('supp_' columns at the end)
proteo.mset <- ProMetIS:::metadata_select(proteo.mset, step.c = "2_post_processed")
Saving
for (set.c in ProMetIS::proteo_sets.vc()) { eset <- proteo.mset[[set.c]] # an 'id' extra column has been automatically created in the sampleMetadata by # 'MultiDataSet' and must be removed before saving pdata.df <- Biobase::pData(eset) pdata.df[, "id"] <- NULL Biobase::pData(eset) <- pdata.df phenomis::writing(eset, file.path(gsub(ProMetIS::data_dir.c(), "../../ProMetIS/inst/extdata", ProMetIS::post_processed_dir.c()), set.c), overwrite.l = TRUE) }
Metabolomics
Loading
metabo.mset <- phenomis::reading(ProMetIS::processed_dir.c(), subsets.vc = ProMetIS::metabo_sets.vc(), output.c = "set", report.c = "none")
Normalization of the dataMatrix
The workflow consists of the following steps:
| step | liver_c18hypersil_pos | liver_hilic_neg | plasma_c18hypersil_pos | plasma_hilic_neg | plasma_c18acquity_pos | plasma_c18acquity_neg | |-----------|-----------|-----------|-----------|-----------|-----------|-----------| | 0.4 \<= RT \<= 22 | | | | | X | X | | blank/samp \<= 0.33 | X | X | X | X | X | X | | poolDil: cor >= 0.7 | X | X | X | X | | | | drift loess, span=1 | none | none | pool | pool | sample | sample | | NA \<= 20% & var > 0 | X | X | X | X | X | X | | pool CV \<= 0.3 | X | X | X | X | X | X | | poolCV/sampCV \<= 1 | X | X | X | X | X | X | | corr, dRT, dmz | X | X | X | X | X | X |
metabo.mset <- ProMetIS:::.metabo_postprocessing(metabo.mset = metabo.mset, drift_correct.c = "prometis")
Formatting and ordering sample names
metabo.mset <- ProMetIS:::.format_metabonames(metabo.mset = metabo.mset, mice_id.df = mice_id.df)
Log2 transformation
metabo.mset <- phenomis::transforming(metabo.mset, report.c = "none")
Re-ordering metadata ('supp_' columns at the end)
metabo.mset <- ProMetIS:::metadata_select(metabo.mset, step.c = "2_post_processed")
Saving
for (set.c in ProMetIS::metabo_sets.vc()) { eset <- metabo.mset[[set.c]] # an 'id' extra column has been automatically created in the sampleMetadata by # 'MultiDataSet' and must be removed before saving pdata.df <- Biobase::pData(eset) pdata.df[, "id"] <- NULL Biobase::pData(eset) <- pdata.df phenomis::writing(eset, file.path(gsub(ProMetIS::data_dir.c(), "../../ProMetIS/inst/extdata", ProMetIS::post_processed_dir.c()), set.c), overwrite.l = TRUE) }
R Package Documentation
Browse R Packages
We want your feedback!
Note that we can't provide technical support on individual packages. You should contact the package authors for that.
Embedding an R snippet on your website
Add the following code to your website.
For more information on customizing the embed code, read Embedding Snippets.