,
Friederike Hoffmann [aut]| Title: | Non-Targeted Fluxomics on High-Resolution Mass-Spectrometry Data |
|---|---|
| Description: | Identifying labeled compounds in a 13C-tracer experiment in non-targeted fashion is a cumbersome process. This package facilitates such type of analyses by providing high level quality control plots, deconvoluting and evaluating spectra and performing a multitude of tests in an automatic fashion. The main idea is to use changing intensity ratios of ion pairs from peak list generated with 'xcms' as candidates and evaluate those against base peak chromatograms and spectra information within the raw measurement data automatically. The functionality is described in Hoffmann et al. (2018) <doi:10.1021/acs.analchem.8b00356>. |
| Authors: | Jan Lisec [aut, cre] ,
Friederike Hoffmann [aut] |
| Maintainer: | Jan Lisec <[email protected]> |
| License: | GPL-3 |
| Version: | 0.62.3 |
| Built: | 2024-11-22 04:21:29 UTC |
| Source: | https://github.com/janlisec/hirestec |
DeconvoluteSpectrum will evaluate a list of xcmsRaw objects at a given time (rt) and potential mass (mz1).
The main purpose is to deconvolute the mass spectrum at rt including mz1.
DeconvoluteSpectrum( dat = NULL, rt = NULL, rt_dev = 3, mz1 = NULL, mz_dev = 0.003, use.mz.adjust = FALSE, ionization = c("APCI", "ESI")[1], smooth = 0 )DeconvoluteSpectrum( dat = NULL, rt = NULL, rt_dev = 3, mz1 = NULL, mz_dev = 0.003, use.mz.adjust = FALSE, ionization = c("APCI", "ESI")[1], smooth = 0 )
dat |
A list of xcmsRaws or an xcmsSet object. |
rt |
Retention time to search for maxima. |
rt_dev |
Allowed retention time window. |
mz1 |
If specified, ensure that this mass is included in the spectrum (assumed base peak). NULL otherwise. |
mz_dev |
Allowed mz deviation [Da]. |
use.mz.adjust |
Will adjust mz on an experiment wide basis. |
ionization |
Either APCI or ESI. Choice will modify some internal parameters and checks performed. |
smooth |
Smoothing parameter passed on to getMultipleBPC. |
Will test all mz at spectrum of base peak within range for co-apex, rt diff and ratio consistency/correlation over a set of samples.
A pseudo spectrum at rt (containing mz1 if specified). Effectively a 2-column matrix (mz, int) with rt as attribute
# Please use examples from previous versions as xcms (and xcms objects) # are no longer supported during CRAN checks leading to package rejection # if included (and I do not know a work around). :(# Please use examples from previous versions as xcms (and xcms objects) # are no longer supported during CRAN checks leading to package rejection # if included (and I do not know a work around). :(
EvaluateCandidateListAgainstRawData will analyze an xcmsSet result for mass pairs (mz1, mz2) with changes due to 13C incorporation.
EvaluateCandidateListAgainstRawData( x = NULL, tp = NULL, gr = NULL, dat = NULL, dmz = 0.025, drt = 1, dEcut = 1, Pcut = 0.01, Icut = 1000, method = c("APCI", "ESI")[1], rolp = c("non", "pos", "neg", "all")[2], smooth = 0 )EvaluateCandidateListAgainstRawData( x = NULL, tp = NULL, gr = NULL, dat = NULL, dmz = 0.025, drt = 1, dEcut = 1, Pcut = 0.01, Icut = 1000, method = c("APCI", "ESI")[1], rolp = c("non", "pos", "neg", "all")[2], smooth = 0 )
x |
Dataframe of results (output of EvaluatePairsFromXCMSet). |
tp |
Timepoint. |
gr |
group, e.g. different genotypes or concentrations. |
dat |
list of xcmsRaw's for deconvolution and plotting. |
dmz |
Allowed mass deviation in Da (for BPC extraction). |
drt |
Allowed rt deviation in seconds (for get extraction). |
dEcut |
Minimum required change in enrichment before a candidate ID is assigned. |
Pcut |
Maximum allowed P value before a candidate ID is assigned. |
Icut |
Minimum required median peak intensity before a candidate ID is assigned. |
method |
Either APCI or ESI. Choice will modify some internal parameters and checks performed. |
rolp |
RemoveOverLappingPeaks paramter. |
smooth |
Smoothing parameter passed to getMultipleBPC. |
This function will evaluate candidate mz pairs found within an xcmsSet object by EvaluatePairsFromXCMSSet against the raw measurement data. A special parameter is 'rolp' which can be set to 'non', 'pos', 'neg' or 'all'. It will influence the time performance of the function be determining how many peaks are effectively tested. If 'rolp' is set to 'non', no overlapping peaks will be skipped, every individual mz-pair will be sequentially evaluated (slow but most informative). If it is set to 'pos' or 'neg', overlapping peaks (determined by experiment wide deconvolution) will not be tested aditionally for positive or negative hits ('neg' is standard). If set to 'all' overlapping peaks will always be removed from the list of mz-pairs to be tested (fast).
A list of evaluation results.
# Please use examples from previous versions as xcms (and xcms objects) # are no longer supported during CRAN checks leading to package rejection # if included (and I do not know a work around). :(# Please use examples from previous versions as xcms (and xcms objects) # are no longer supported during CRAN checks leading to package rejection # if included (and I do not know a work around). :(
EvaluatePairsFromXCMSSet will analyze an xcmsSet result for mass pairs (mz1, mz2) with changes due to any 13C incorporation.
EvaluatePairsFromXCMSSet( xg = NULL, tp = NULL, gr = NULL, drt = 1, dmz = 0.025, mz_iso = 1.00335, n = 6, method = c("APCI", "ESI")[1], specific_row = NULL, testing = FALSE, silent = FALSE )EvaluatePairsFromXCMSSet( xg = NULL, tp = NULL, gr = NULL, drt = 1, dmz = 0.025, mz_iso = 1.00335, n = 6, method = c("APCI", "ESI")[1], specific_row = NULL, testing = FALSE, silent = FALSE )
xg |
xcmsSet object with group information. |
tp |
Timepoint information for all samples (obviously required, internally converted to factor). |
gr |
Group information for all samples, e.g. different genotypes or concentrations (optional, factor). |
drt |
Allowed rt deviation in time units of xcmsSet (usually seconds) to test for candidates. |
dmz |
Allowed mass deviation in Da. |
mz_iso |
Mass defect of the isotope under investigation. |
n |
Number of maximal incorporated carbons to test. |
method |
Currently APCI or ESI. If APCI, dmz will be modified depending on n (see details). |
specific_row |
A single row from groupval(xg) to process. |
testing |
Stop in function using browser() if specific_row is specified; can be a isotope number, i.e. 3 will stop at third isotope. |
silent |
Suppress warnings and console output if TRUE. |
Using 'APCI' as method assumes that (i) you analyze TMS-derivatized compounds and (ii) your MS resolution does not allow to seperate Si and C isotopes but reportes an intermediate mass as m/z. In this case you will find carbon isotopes below there expected masses, i.e. M+1 would be 1.001mDa apart from M+0 instead of 1.003. The effect is increased with isotope number, i.e. M+6 will be ~20mDa below the expected value. Hence, selecting method 'APCI' will combine your selected dmz with a allowed deviation due to Si-isotope caused mass shifts. Use 'ESI' if you are not sure if this effect takes place in your settings.
A dataframe with all observable pairs within the provided xcmsSet peak list including mean group intensities and P values.
# Please use examples from previous versions as xcms (and xcms objects) are # no longer supported during CRAN checks leading to package rejection # if included (and I do not know a work around). ## Not run: load(xcms_cand) head(xcms_cand[order(xcms_cand$P), ]) ## End(Not run)# Please use examples from previous versions as xcms (and xcms objects) are # no longer supported during CRAN checks leading to package rejection # if included (and I do not know a work around). ## Not run: load(xcms_cand) head(xcms_cand[order(xcms_cand$P), ]) ## End(Not run)
GenerateCandXLSX will produce a XLSX of a list containing test results objects.
GenerateCandXLSX(res_list = NULL, xlsx_file = NULL, rejected = FALSE)GenerateCandXLSX(res_list = NULL, xlsx_file = NULL, rejected = FALSE)
res_list |
A list of result objects (each testing an individual mz pair). |
xlsx_file |
File name. |
rejected |
Logical. Return rejected if TRUE. |
Not yet.
Candidate table as data.frame.
# load evaluation result of example data data(res_list) # generate table within R (use xlsx_file to write to file) str(GenerateCandXLSX(res_list)) GenerateCandXLSX(res_list)[, 1:5]# load evaluation result of example data data(res_list) # generate table within R (use xlsx_file to write to file) str(GenerateCandXLSX(res_list)) GenerateCandXLSX(res_list)[, 1:5]
GenerateQCPlots will produce QC plots for a list containing test results objects.
GenerateQCPlots( res_list = NULL, pdf_file = NULL, mfrow = NULL, skip_plots = NULL )GenerateQCPlots( res_list = NULL, pdf_file = NULL, mfrow = NULL, skip_plots = NULL )
res_list |
A list of result objects (each testing an individual mz pair). |
pdf_file |
Either APCI or ESI. Choice will modify some internal parameters and checks performed. |
mfrow |
If NULL automatically determined, otherwise useful to specify a layout. |
skip_plots |
NULL or numeric vector in which case plots with numbers in skip_plots will be empty. |
For individual candidates screen output is reasonable, otherwise a target PDF file should be specified.
NULL.
# load evaluation result of example data data(res_list) # generate Figures on screen (use PDF output for mass evaluation) GenerateQCPlots(res_list[1])# load evaluation result of example data data(res_list) # generate Figures on screen (use PDF output for mass evaluation) GenerateQCPlots(res_list[1])
getMultipleBPC will extract multiple BPCs from an xcmsRaw
object for a vector of mz within the limits given by rt, rt_dev and mz_dev.
getMultipleBPC( x, mz = NULL, mz_dev = 0.005, rt = NULL, rt_dev = 2, zeroVal = NA, smooth = 0, returnEIC = FALSE )getMultipleBPC( x, mz = NULL, mz_dev = 0.005, rt = NULL, rt_dev = 2, zeroVal = NA, smooth = 0, returnEIC = FALSE )
x |
xcmsRaw object. |
mz |
mass vector or NULL (default) to return the TIC. |
mz_dev |
allowed deviations (can be a single numeric, a vector, a matrix with one row (lower bound, upper bound) or a matrix with |
rt |
target time point or NULL (default) to use full scan time. |
rt_dev |
allowed window. |
zeroVal |
Set values <=0 to NA or keep as is with NULL. |
smooth |
Window size for moving average smoother, 0 = no smoothing. |
returnEIC |
Return EIC instead of BPC? |
While there are other functions to extract BPC information from raw data, this one is particularly useful to get all traces belonging to a isotopologue group. It will attach several derived values to the results object, i.e. describing the observed mass shift (deviation from expected value) which is helpful in QC for non-targeted tracer analyses.
A matrix with scan wise (rows) intensities for all requested masses (columns) as either EIC or BPC.
Uses C code modified from XCMS (see citation("xcms")).
# see \link{plotBPC} for an example# see \link{plotBPC} for an example
This is a list defining windows for high res APCI or ESI instrumentation..
data(mz_shift_corrector)data(mz_shift_corrector)
An object of class list of length 3.
plotBPC will plot for each item of a list of result-ojects from getMultipleBPC the BPC traces and the spectrum at the scan where the summed intensity of all ions is max.
plotBPC( bpc = NULL, mfrow = NULL, skip_plots = NULL, ylim = NULL, col = NULL, ids = NULL, type = "both", ann = c("mdev", "mz", "none") )plotBPC( bpc = NULL, mfrow = NULL, skip_plots = NULL, ylim = NULL, col = NULL, ids = NULL, type = "both", ann = c("mdev", "mz", "none") )
bpc |
A bpc object (list of intensity matrixes, rt x mz, including several attributes as attached by getMultipleBPC). |
mfrow |
Specify mfrow explicitely (is optimized internally if NULL to cover n=length(bpc)). |
skip_plots |
Allows to block certain subplots in the mfrow matrix to bettern align replicates. |
ylim |
Can be specified specifically, will be adjusted to overall min/max otherwise. |
col |
Specific color vector for masses may be provided. |
ids |
Specific plot ids may be explicitely provided. |
type |
Switch between co-plot of BPC and Spectrum ("both") or BPC alone ("bpc"). |
ann |
Select value to annotate peaks in spectrum. Usually the mass deviation from the expected value in mDa. |
not yet
A plot to the graphics device and NULL as invisible.
# load example raw data data(res_list) plotBPC(bpc = res_list[[1]][["bpc"]][c(1:2, 13:14)]) plotBPC(bpc = res_list[[1]][["bpc"]][c(1:2, 13:14)], ann="mz")# load example raw data data(res_list) plotBPC(bpc = res_list[[1]][["bpc"]][c(1:2, 13:14)]) plotBPC(bpc = res_list[[1]][["bpc"]][c(1:2, 13:14)], ann="mz")
plotMID will plot a Mass Isotopomer Distribution (MID).
plotMID( mid = NULL, gr = NULL, name = "unknown", contr = NULL, stackedbars = FALSE, subplot_ylim = 100, ... )plotMID( mid = NULL, gr = NULL, name = "unknown", contr = NULL, stackedbars = FALSE, subplot_ylim = 100, ... )
mid |
Matrix of corrected mass isotopomer distributions. Samples in columns, MID values in rows. |
gr |
Groups, a factor vector of length ncol(mid). |
name |
Name of metabolite. |
contr |
Contrasts. Not yet clear if useful. |
stackedbars |
Alternative plotting layout using stacked bar plot. |
subplot_ylim |
Calculate ylim individually per subplot if 0, show full range in all subplots if 100 and limit to the minimal specified number otherwise. |
... |
Further arguments to 'boxplot' or 'barplot' (depending on 'stackedbars'). |
Not yet.
A plot.
mid <- matrix(c(seq(0, 0.3, 0.1), seq(1, 0.7, -0.1)), byrow = TRUE, nrow = 2) gr <- gl(2, 2, labels = letters[1:2]) plotMID(mid = mid, gr = gr, name = "Metabolite X") plotMID(mid = mid, gr = gr, stackedbars = TRUE, las = 1, xlab = "MID", legend.text = c("x", "y")) lt <- paste0("M", 0:1) rownames(mid) <- lt plotMID(mid = mid, gr = gr, stackedbars = TRUE, las = 1, xlab = "MID", legend.text = lt) plotMID(mid = mid[, 2, drop = FALSE], stackedbars = TRUE, col = c(3, 4)) colnames(mid) <- paste0("S", 1:4) gr2 <- gl(n = 1, k = 1, labels = "bla") plotMID(mid = mid[, 2, drop = FALSE], gr = gr2, stackedbars = TRUE, name = NULL) plotMID(mid = mid, gr = factor(colnames(mid)), stackedbars = TRUE, name = NULL)mid <- matrix(c(seq(0, 0.3, 0.1), seq(1, 0.7, -0.1)), byrow = TRUE, nrow = 2) gr <- gl(2, 2, labels = letters[1:2]) plotMID(mid = mid, gr = gr, name = "Metabolite X") plotMID(mid = mid, gr = gr, stackedbars = TRUE, las = 1, xlab = "MID", legend.text = c("x", "y")) lt <- paste0("M", 0:1) rownames(mid) <- lt plotMID(mid = mid, gr = gr, stackedbars = TRUE, las = 1, xlab = "MID", legend.text = lt) plotMID(mid = mid[, 2, drop = FALSE], stackedbars = TRUE, col = c(3, 4)) colnames(mid) <- paste0("S", 1:4) gr2 <- gl(n = 1, k = 1, labels = "bla") plotMID(mid = mid[, 2, drop = FALSE], gr = gr2, stackedbars = TRUE, name = NULL) plotMID(mid = mid, gr = factor(colnames(mid)), stackedbars = TRUE, name = NULL)
This is a list containing the evaluations results established based on processing example data with EvaluateCandidateListAgainstRawData.
data(res_list)data(res_list)
An object of class list of length 14.
This data frame contains the sample definition of 18 samples from a larger experiment.
data(sam)data(sam)
An object of class data.frame with 18 rows and 8 columns.
This data frame contains the analysis result of an xcmsSet which can not be provided via CRAN anymore using EvaluatePairsFromXCMSSet with respect to interesting m/z-pairs.
data(xcms_cand)data(xcms_cand)
An object of class data.frame with 72 rows and 19 columns.