,
Fayyaz Afsar [aut],
Steffen Neumann [aut],
Asa Ben-Hur [aut],
Jessica Prenni [aut],
Helge Hecht [cre] | Title: | Mass Spectrometry Metabolomics Feature Clustering and Interpretation |
|---|---|
| Description: | A feature clustering algorithm for non-targeted mass spectrometric metabolomics data. This method is compatible with gas and liquid chromatography coupled mass spectrometry, including indiscriminant tandem mass spectrometry <DOI: 10.1021/ac501530d> data. |
| Authors: | Corey D. Broeckling [aut] ,
Fayyaz Afsar [aut],
Steffen Neumann [aut],
Asa Ben-Hur [aut],
Jessica Prenni [aut],
Helge Hecht [cre] |
| Maintainer: | Helge Hecht <[email protected]> |
| License: | MIT |
| Version: | 1.3.0 |
| Built: | 2024-10-27 06:21:57 UTC |
| Source: | https://github.com/cbroeckl/ramclustr |
add rc.feature.replace.na params in ramclustObj
add_params(ramclustObj, params, param_name)add_params(ramclustObj, params, param_name)
ramclustObj |
ramclustObj containing MSdata with optional MSMSdata (MSe, DIA, idMSMS) |
params |
vector containing parameters to add |
param_name |
name of the parameter/step |
ramclustR object with rc.feature.replace.na params added.
check provided arguments
check_arguments_filter.blanks(ramclustObj, sn)check_arguments_filter.blanks(ramclustObj, sn)
ramclustObj |
ramclustObj containing MSdata with optional MSMSdata (MSe, DIA, idMSMS) |
sn |
numeric defines the ratio for 'signal'. i.e. sn = 3 indicates that signal intensity must be 3 fold higher in sample than in blanks, on average, to be retained. |
check provided arguments
check_arguments_filter.cv(ramclustObj, qc.tag)check_arguments_filter.cv(ramclustObj, qc.tag)
ramclustObj |
ramclustObj containing MSdata with optional MSMSdata (MSe, DIA, idMSMS) |
qc.tag |
character vector of length one or two. If length is two, enter search string and factor name in $phenoData slot (i.e. c("QC", "sample.type"). If length one (i.e. "QC"), will search for this string in the 'sample.names' slot by default. |
check provided arguments
check_arguments_replace.na( ramclustObj, replace.int, replace.noise, replace.zero )check_arguments_replace.na( ramclustObj, replace.int, replace.noise, replace.zero )
ramclustObj |
ramclustObj containing MSdata with optional MSMSdata (MSe, DIA, idMSMS) |
replace.int |
default = 0.1. proportion of minimum feature value to replace NA (or zero) values with |
replace.noise |
default = 0.1. proportion ofreplace.int value by which noise is added via 'jitter' |
replace.zero |
logical if TRUE, any zero values are replaced with noise as if they were NA values |
check if MS data contains mz and rt, and if MSMS data is present feature names and sample names are identical
checks( ms1_featureDefinitions = NULL, ms1_featureValues = NULL, ms2_featureValues = NULL, feature_names = NULL )checks( ms1_featureDefinitions = NULL, ms1_featureValues = NULL, ms2_featureValues = NULL, feature_names = NULL )
ms1_featureDefinitions |
dataframe with metadata with columns: mz, rt, feature names containing MS data |
ms1_featureValues |
dataframe with rownames = sample names, colnames = feature names containing MS data |
ms2_featureValues |
dataframe with rownames = sample names, colnames = feature names containing MSMS data |
feature_names |
feature names extracted from the data |
compute data frame to use in ramclustObj
compute_do.sets(ramclustObj)compute_do.sets(ramclustObj)
ramclustObj |
ramclustObj containing MSdata with optional MSMSdata (MSe, DIA, idMSMS) |
vector which is used to select data frame to use in ramclustObj
further aggregate by sample names for 'SpecAbundAve' dataset
compute_SpecAbundAve(ramclustObj = NULL)compute_SpecAbundAve(ramclustObj = NULL)
ramclustObj |
ramclustObj containing MSdata with optional MSMSdata (MSe, DIA, idMSMS) |
ramclustR object with aggregate by sample names for 'SpecAbundAve' dataset
compute weighted.mean intensity of feature in ms/msms level data
compute_wt_mean(data, global.min, fmz, ensure.no.na)compute_wt_mean(data, global.min, fmz, ensure.no.na)
data |
feature in ms/msms level data |
global.min |
minimum intensity in ms/msms level data |
fmz |
feature retention time |
ensure.no.na |
logical: if TRUE, any 'NA' values in msint and/or msmsint are replaced with numerical values based on 10 percent of feature min plus noise. Used to ensure that spectra are not written with NA values. |
weighted.mean intensity of feature in ms/msms level data
create ramclustr Object
create_ramclustObj( ExpDes = NULL, input_history = NULL, MSdata = NULL, MSMSdata = NULL, frt = NULL, fmz = NULL, st = NULL, phenoData = NULL, feature_names = NULL, sample_names = NULL, xcmsOrd = NULL, ensure.no.na = TRUE )create_ramclustObj( ExpDes = NULL, input_history = NULL, MSdata = NULL, MSMSdata = NULL, frt = NULL, fmz = NULL, st = NULL, phenoData = NULL, feature_names = NULL, sample_names = NULL, xcmsOrd = NULL, ensure.no.na = TRUE )
ExpDes |
either an R object created by R ExpDes object: data used for record keeping and labelling msp spectral output |
input_history |
input history |
MSdata |
dataframe containing MS Data |
MSMSdata |
dataframe containing MSMS Data |
frt |
feature retention time, in whatever units were fed in |
fmz |
feature retention time |
st |
numeric: sigma t - time similarity decay value |
phenoData |
dataframe containing phenoData |
feature_names |
feature names extracted from the data |
sample_names |
sample names extracted from the data |
xcmsOrd |
original xcms order of features, for back-referencing when necessary |
ensure.no.na |
logical: if TRUE, any 'NA' values in msint and/or msmsint are replaced with numerical values based on 10 percent of feature min plus noise. Used to ensure that spectra are not written with NA values. |
an ramclustR object. this object is formatted as an hclust object with additional slots for holding feature and compound data.
define samples in each set
define_samples(ramclustObj, tag, return.logical = FALSE)define_samples(ramclustObj, tag, return.logical = FALSE)
ramclustObj |
ramclustObj containing MSdata with optional MSMSdata (MSe, DIA, idMSMS) |
tag |
character vector of length one or two. If length is two, enter search string and factor name in $phenoData slot (i.e. c("QC", "sample.type"). If length one (i.e. "QC"), will search for this string in the 'sample.names' slot by default. |
logical |
optionally convert numeric vector with length equal to the number of matched samples to a logical vector of length equal to number of samples, with TRUE representing matching samples. |
samples found using the tag
Create an Experimental Design R object for record-keeping and msp output
defineExperiment(csv = FALSE, force.skip = FALSE)defineExperiment(csv = FALSE, force.skip = FALSE)
csv |
logical or filepath. If csv = TRUE , csv template called "ExpDes.csv" will be written to your working directory. you will fill this in manually, ensuring that when you save you retain csv format. ramclustR will then read this file in and and format appropriately. If csv = FALSE, a pop up window will appear (in windows, at least) asking for input. If a character string with full path (and file name) to a csv file is given, this will allow you to read in a previously edited csv file. |
force.skip |
logical. If TRUE, ramclustR creates a pseudo-filled ExpDes object to enable testing of functionality. Not recommended for real data, as your exported spectra will be improperly labelled. |
an Exp Des R object which will be used for record keeping and writing spectra data.
Corey Broeckling
Broeckling CD, Afsar FA, Neumann S, Ben-Hur A, Prenni JE. RAMClust: a novel feature clustering method enables spectral-matching-based annotation for metabolomics data. Anal Chem. 2014 Jul 15;86(14):6812-7. doi: 10.1021/ac501530d. Epub 2014 Jun 26. PubMed PMID: 24927477.
Broeckling CD, Ganna A, Layer M, Brown K, Sutton B, Ingelsson E, Peers G, Prenni JE. Enabling Efficient and Confident Annotation of LC-MS Metabolomics Data through MS1 Spectrum and Time Prediction. Anal Chem. 2016 Sep 20;88(18):9226-34. doi: 10.1021/acs.analchem.6b02479. Epub 2016 Sep 8. PubMed PMID: 7560453.
Cluster annotation function: inference of 'M' - molecular weight of the compound giving rise to each spectrum - using the InterpretMSSpectrum::findMain function
doFindmain( ramclustObj = NULL, cmpd = NULL, mode = "positive", mzabs.error = 0.005, ppm.error = 10, ads = NULL, nls = NULL, scoring = "auto", plot.findmain = TRUE, writeMat = TRUE, writeMS = TRUE, use.z = TRUE )doFindmain( ramclustObj = NULL, cmpd = NULL, mode = "positive", mzabs.error = 0.005, ppm.error = 10, ads = NULL, nls = NULL, scoring = "auto", plot.findmain = TRUE, writeMat = TRUE, writeMS = TRUE, use.z = TRUE )
ramclustObj |
ramclustR object to annotate. |
cmpd |
integer: vector defining compound numbers to annotated. if NULL (default), all compounds |
mode |
character: "positive" or "negative" |
mzabs.error |
numeric: absolute mass deviation allowd, default = 0.01 |
ppm.error |
numeric: ppm mass error _added_ to mzabs.error, default = 10 |
ads |
character: vector of allowed adducts, i.e. c("[M+H]+"). if NULL, default positive mode values of H+, Na+, K+, and NH4+, as monomer, dimer, and trimer, are assigned. Negative mode include "[M-H]-", "[M+Na-2H]-", "[M+K-2H]-", "[M+CH2O2-H]-" as monomer, dimer, and trimer. |
nls |
character: vector of allowed neutral losses, i.e. c("[M+H-H2O]+"). if NULL, an extensive list derived from CAMERA's will be used. |
scoring |
character: one of 'imss' , 'ramclustr', or 'auto'. default = 'auto'. see details. |
plot.findmain |
logical: should pdf polts be generated for evaluation? detfault = TRUE. PDF saved to working.directory/spectra |
writeMat |
logical: should individual .mat files (for MSFinder) be generated in a 'mat' subdirectory in the 'spectra' folder? default = TRUE. |
writeMS |
logical: should individual .ms files (for Sirius) be generated in a 'ms' subdirectory in the 'spectra' folder? default = TRUE. Note that no import functions are yet written for Sirius output. |
use.z |
logical: if you have previously run the 'assign.z' function from ramclustR, there will be a slot reflecting the feature mass after accounting for charge (fm) - if TRUE this is used instead of feature m/z (fmz) in interpreting MS data and exporting spectra for annotation. |
a partially annotated ramclustR object. base structure is that of a standard R heirarchical clustering output, with additional slots described in ramclustR documentation (?ramclustR). New slots added after using the interpretMSSpectrum functionality include those described below.
$M: The inferred molecular weight of the compound giving rise to the each spectrum
$M.ppm: The ppm error of all the MS signals annotated, high error values should be considered 'red flags'.
$M.ann: The annotated spectrum supporting the interpretation of M
$use.findmain: Logical vector indicating whether findmain scoring (TRUE) or ramclustR scoring (FALSE) was used to support inference of M. By default, findmain scoring is used. When ramclustR scoring differs from findmain scoring, the scoring metric which predicts higher M is selected.
$M.ramclustr: M selected using ramclustR scoring
$M.ppm.ramclustr: ppm error of M selected using ramclustR scoring. Used to resolve concflicts between ramclustR and findmain M assignment when scoring = auto.
$M.ann.ramclustr: annotated spectrum supporting M using ramclustR scoring
$M.nann.ramclustr: number of masses annotated using ramclustR scoring. Used to resolve concflicts between ramclustR and findmain M assignment when scoring = auto.
$M.space.ramclustr: the 'space' of scores between the best and second best ramclustR scores. Calculated as a ratio. Used to resolve concflicts between ramclustR and findmain M assignment when scoring = auto.
$M.findmain: M selected using findmain scoring
$M.ppm.findmain: ppm error of M selected using findmain scoring. Used to resolve concflicts between ramclustR and findmain M assignment when scoring = auto.
$M.ann.findmain: annotated spectrum supporting M using findmain scoring
$M.nann.findmain: number of masses annotated using findmain scoring. Used to resolve concflicts between ramclustR and findmain M assignment when scoring = auto.
$M.space.findmain: the 'space' of scores between the best and second best findmain scores. Calculated as a ratio. Used to resolve concflicts between ramclustR and findmain M assignment when scoring = auto.
Corey Broeckling
Jaeger C, ... Lisec J. Compound annotation in liquid chromatography/high-resolution mass spectrometry based metabolomics: robust adduct ion determination as a prerequisite to structure prediction in electrospray ionization mass spectra. Rapid Commun Mass Spectrom. 2017 Aug 15;31(15):1261-1266. doi: 10.1002/rcm.7905. PubMed PMID: 28499062.
Broeckling CD, Afsar FA, Neumann S, Ben-Hur A, Prenni JE. RAMClust: a novel feature clustering method enables spectral-matching-based annotation for metabolomics data. Anal Chem. 2014 Jul 15;86(14):6812-7. doi: 10.1021/ac501530d. Epub 2014 Jun 26. PubMed PMID: 24927477.
Broeckling CD, Ganna A, Layer M, Brown K, Sutton B, Ingelsson E, Peers G, Prenni JE. Enabling Efficient and Confident Annotation of LC-MS Metabolomics Data through MS1 Spectrum and Time Prediction. Anal Chem. 2016 Sep 20;88(18):9226-34. doi: 10.1021/acs.analchem.6b02479. Epub 2016 Sep 8. PubMed PMID: 7560453.
export one of 'SpecAbund', 'SpecAbundAve', 'MSdata' or 'MSMSdata' from an RC object to csv
exportDataset( ramclustObj = NULL, which.data = "SpecAbund", label.by = "ann", appendFactors = TRUE )exportDataset( ramclustObj = NULL, which.data = "SpecAbund", label.by = "ann", appendFactors = TRUE )
ramclustObj |
ramclustR object to export from |
which.data |
name of dataset to export. SpecAbund, SpecAbundAve, MSdata, or MSMSdata |
label.by |
either 'ann' or 'cmpd', generally. name of ramclustObj slot used as csv header for each column (compound) |
appendFactors |
logical. If TRUE (default) the factor data frame is appended to the left side of the dataset. |
Useful for exporting the processed signal intensity matrix to csv for analysis elsewhere.
nothing is returned. file exported as csf to 'datasets/*.csv'
Corey Broeckling
Broeckling CD, Afsar FA, Neumann S, Ben-Hur A, Prenni JE. RAMClust: a novel feature clustering method enables spectral-matching-based annotation for metabolomics data. Anal Chem. 2014 Jul 15;86(14):6812-7. doi: 10.1021/ac501530d. Epub 2014 Jun 26. PubMed PMID: 24927477.
filter blanks
filter_blanks(ramclustObj, keep, d1)filter_blanks(ramclustObj, keep, d1)
ramclustObj |
ramclustObj containing MSdata with optional MSMSdata (MSe, DIA, idMSMS) |
keep |
union of which signal is at least 3x larger, output of filter_signal() |
d1 |
MS Data |
ramclustObj object with feature.filter.blanks
filter to keep only 'good' features
filter_good_features(ramclustObj, keep)filter_good_features(ramclustObj, keep)
ramclustObj |
ramclustObj containing MSdata with optional MSMSdata (MSe, DIA, idMSMS) |
keep |
features to keep. output of find_good_features(). |
ramclustR object filtered to keep only 'good' features
filter signal
filter_signal(ms.qc.mean, ms.blank.mean, sn)filter_signal(ms.qc.mean, ms.blank.mean, sn)
ms.qc.mean |
ms qc mean signal intensities |
ms.blank.mean |
ms blank mean signal intensities |
sn |
numeric defines the ratio for 'signal'. i.e. sn = 3 indicates that signal intensity must be 3 fold higher in sample than in blanks, on average, to be retained. |
union of which signal is at least 3x larger
find 'good' features, acceptable CV at either MS or MSMS level results in keeping
find_good_features(ramclustObj, do.sets, max.cv, qc)find_good_features(ramclustObj, do.sets, max.cv, qc)
ramclustObj |
ramclustObj containing MSdata with optional MSMSdata (MSe, DIA, idMSMS) |
do.sets |
select data frame to use. |
max.cv |
numeric maximum allowable cv for any feature. default = 0.5 |
qc |
QC samples found by define_samples |
ramclustR object
features to keep
see if any features match a given mass, and whether they are plausibly M0
findfeature( ramclustObj = NULL, mz = NULL, mztol = 0.02, rt = NULL, rttol = 2, iso.rttol = 2, zmax = 6, m.check = TRUE )findfeature( ramclustObj = NULL, mz = NULL, mztol = 0.02, rt = NULL, rttol = 2, iso.rttol = 2, zmax = 6, m.check = TRUE )
ramclustObj |
R object: the ramclustR object to explore |
mz |
numeric: mz value to search for |
mztol |
numeric: absolute mass tolerance around mz |
rt |
numeric: optional rt value to search for (generally in seconds, though use whatever units your data is in) |
rttol |
numeric: absolute retention time tolerance around rt. |
iso.rttol |
numeric: when examining isotope patterns, feature retention time tolerance around features matching mz +- mztol |
zmax |
integer: maximum charge state to consider. default is 6. |
m.check |
logical: check whether the matching masses are plausibly M0. That is, we look for ions 1 proton mass (from charge state 1:zmax) below the target m/z at the same time that have intensities consistent with target ion being a non-M0 isotope. |
a convenience function to perform a targeted search of all features for a mass of interest. Also performs a crude plausibility check as to whether the matched feature could be M0, based on the assumption of approximately 1 carbon per 17 m/z units and natural isotopic abundance of 1.1
returns a table to the console listing masses which match, their retention time and intensity, and whether it appears to be plausible as M0
Corey Broeckling
see if any features match a given mass, and whether they are plausibly M0
findmass( ramclustObj = NULL, mz = NULL, mztol = 0.02, rttol = 2, zmax = 6, m.check = TRUE )findmass( ramclustObj = NULL, mz = NULL, mztol = 0.02, rttol = 2, zmax = 6, m.check = TRUE )
ramclustObj |
R object: the ramclustR object to explore |
mz |
numeric: mz value to search for |
mztol |
numeric: absolute mass tolerance around mz |
rttol |
numeric: when examining isotope patterns, feature retention time tolerance around features matching mz +- mztol |
zmax |
integer: maximum charge state to consider. default is 6. |
m.check |
logical: check whether the matching masses are plausibly M0. That is, we look for ions 1 proton mass (from charge state 1:zmax) below the target m/z at the same time that have intensities consistent with target ion being a non-M0 isotope. |
a convenience function to perform a targeted search of all feaures for a mass of interest. Also performs a crude plausibility check as to whether the matched feature could be M0, based on the assumption of approximately 1 carbon per 17 m/z units and natural isotopic abundance of 1.1
returns a table to the console listing masses which match, their retention time and intensity, and whether it appears to be plausible as M0
Corey Broeckling
get Experimental Design
get_ExpDes(csv.in)get_ExpDes(csv.in)
csv.in |
Experimental Design read from csv |
list containing design and instrument
get instrument platform
get_instrument_platform(design)get_instrument_platform(design)
design |
data frame containing Experimental Design |
instrument platform
retrieve and parse sample names, retrieve metabolite data. returns as list of two data frames
getData( ramclustObj = NULL, which.data = "SpecAbund", delim = "-", cmpdlabel = "cmpd", filter = FALSE )getData( ramclustObj = NULL, which.data = "SpecAbund", delim = "-", cmpdlabel = "cmpd", filter = FALSE )
ramclustObj |
ramclustR object to retrieve data from |
which.data |
character; which dataset (SpecAbund or SpecAbundAve) to reference |
delim |
character; "-" by default - the delimiter for parsing sample names to factors |
cmpdlabel |
= "cmpd"; label the data with the annotation. can also be set to 'ann' for column names assigned as annotatins. |
filter |
= TRUE; logical, if TRUE, checks for $cmpd.use slot generated by rc.cmpd.cv.filter() function, and only gets acceptable compounds. |
convenience function for parsing sample names and returning a dataset.
returns a list of length 3: $design is the experimental sample factors after parsing by the delim, $data is the dataset, $full.data is merged $des and $data data.frames.
Corey Broeckling
calculate MS mean signal intensities
mean_signal_intensities(data, sample)mean_signal_intensities(data, sample)
data |
MS/MSMS data |
sample |
sample found using the tag, output of define_samples() |
mean signal intensities
merge two ramclustR objects
mergeRCobjects( ramclustObj.1 = NULL, ramclustObj.2 = NULL, mztol = 0.02, rttol = 30, course.rt.adj = NULL, mzwt = 2, rtwt = 1, intwt = 3 )mergeRCobjects( ramclustObj.1 = NULL, ramclustObj.2 = NULL, mztol = 0.02, rttol = 30, course.rt.adj = NULL, mzwt = 2, rtwt = 1, intwt = 3 )
ramclustObj.1 |
ramclustR object 1: this object will be the base for the new object. That is all the features from ramclustObj.1 will be retained. |
ramclustObj.2 |
ramclustR object 2: this object will mapped and appended to racmlustObj1. That is only features which appear consistent with those from ramclustObj.1 will be retained. |
mztol |
numeric: absolute mass tolerance around mz |
rttol |
numeric: feature retention time tolerance. Value set by this option will be used during the initial anchor mapping phase. Two times the standard error of the rt loess correction will be used for the full mapping. |
course.rt.adj |
numeric: default = NULL. optional approximate retention time shift between ramclustObj.1 and ramclustObj.2. i.e if the retention time of ramclustObj.1 is on average 15 seconds longer than that of ramclustobj.2, enter '15'. if 1 is less than 2, enter a negative number. This is applied before mapping to enable a smaller 'rttol' value to be used. |
mzwt |
numeric: when mapping features, weighting value used for similarities between feature mass values (see rtwt, intwt) |
rtwt |
numeric: when mapping features, weighting value used for similarities between feature retention time values (see mzwt, intwt) |
intwt |
numeric: when mapping features, weighting value used for similarities between ranked signal intensity values (see rtwt, mzwt) |
Two ramclustR objects are merged with this function, mapping features between them. The first (ramclustObj.1) object use used as the template - all data in it is retained. ramclustObj.2 is mapped to ramclustObj.1 feature by feature - only mapped features are retained. A new ramlcustObj is returned, with a new SpecAbund dataset with the same column number as the ramclustObj.1$SpecAbund set.
returns a ramclustR object. All values from ramclustObj.1 are retained. SpecAbund dataset from ramclustObj.1 is moved to RC$SpecAbund.1, where RC is the new ramclustObj.
Corey Broeckling
normalize data using batch.qc
normalized_data_batch_qc( data = NULL, batch = NULL, order = NULL, qc = NULL, qc.inj.range = 20 )normalized_data_batch_qc( data = NULL, batch = NULL, order = NULL, qc = NULL, qc.inj.range = 20 )
data |
feature in ms/msms level data |
batch |
integer vector with length equal to number of injections in xset or csv file or dataframe |
order |
integer vector with length equal to number of injections in xset or csv file or dataframe |
qc |
logical vector with length equal to number of injections in xset or csv file or dataframe |
qc.inj.range |
integer: how many injections around each injection are to be scanned for presence of QC samples when using batch.qc normalization? A good rule of thumb is between 1 and 3 times the typical injection span between QC injections. i.e. if you inject QC ever 7 samples, set this to between 7 and 21. smaller values provide more local precision but make normalization sensitive to individual poor outliers (though these are first removed using the boxplot function outlier detection), while wider values provide less local precision in normalization but better stability to individual peak areas. |
normalized data.
normalize data using TIC
normalized_data_tic(ramclustObj = NULL)normalized_data_tic(ramclustObj = NULL)
ramclustObj |
ramclustObj containing MSdata with optional MSMSdata (MSe, DIA, idMSMS) |
ramclustR object with total extracted ion normalized data.
order the datasets first by batch and run order
order_datasets(order = NULL, batch = NULL, qc = NULL, data = NULL)order_datasets(order = NULL, batch = NULL, qc = NULL, data = NULL)
order |
integer vector with length equal to number of injections in xset or csv file or dataframe |
batch |
integer vector with length equal to number of injections in xset or csv file or dataframe |
qc |
logical vector with length equal to number of injections in xset or csv file or dataframe |
data |
feature in ms/msms level data |
ordered feature in ms/msms level data, order, batch, qc
Main clustering function for grouping features based on their analytical behavior.
ramclustR( xcmsObj = NULL, ms = NULL, pheno_csv = NULL, idmsms = NULL, taglocation = "filepaths", MStag = NULL, idMSMStag = NULL, featdelim = "_", timepos = 2, st = NULL, sr = NULL, maxt = NULL, deepSplit = FALSE, blocksize = 2000, mult = 5, hmax = NULL, sampNameCol = 1, collapse = TRUE, usePheno = TRUE, mspout = TRUE, ExpDes = NULL, normalize = "TIC", qc.inj.range = 20, order = NULL, batch = NULL, qc = NULL, minModuleSize = 2, linkage = "average", mzdec = 3, cor.method = "pearson", rt.only.low.n = TRUE, replace.zeros = TRUE )ramclustR( xcmsObj = NULL, ms = NULL, pheno_csv = NULL, idmsms = NULL, taglocation = "filepaths", MStag = NULL, idMSMStag = NULL, featdelim = "_", timepos = 2, st = NULL, sr = NULL, maxt = NULL, deepSplit = FALSE, blocksize = 2000, mult = 5, hmax = NULL, sampNameCol = 1, collapse = TRUE, usePheno = TRUE, mspout = TRUE, ExpDes = NULL, normalize = "TIC", qc.inj.range = 20, order = NULL, batch = NULL, qc = NULL, minModuleSize = 2, linkage = "average", mzdec = 3, cor.method = "pearson", rt.only.low.n = TRUE, replace.zeros = TRUE )
xcmsObj |
xcmsObject: containing grouped feature data for clustering by ramclustR |
ms |
filepath: optional csv input. Features as columns, rows as samples. Column header mz_rt |
pheno_csv |
filepath: optional csv input containing phenoData |
idmsms |
filepath: optional idMSMS / MSe csv data. same dim and names as ms required |
taglocation |
character: "filepaths" by default, "phenoData[,1]" is another option. refers to xcms slot |
MStag |
character: character string in 'taglocation' to designat MS / MSe files e.g. "01.cdf" |
idMSMStag |
character: character string in 'taglocation' to designat idMSMS / MSe files e.g. "02.cdf" |
featdelim |
character: how feature mz and rt are delimited in csv import column header e.g. ="-" |
timepos |
integer: which position in delimited column header represents the retention time (csv only) |
st |
numeric: sigma t - time similarity decay value |
sr |
numeric: sigma r - correlational similarity decay value |
maxt |
numeric: maximum time difference to calculate retention similarity for - all values beyond this are assigned similarity of zero |
deepSplit |
logical: controls how agressively the HCA tree is cut - see ?cutreeDynamicTree |
blocksize |
integer: number of features (scans?) processed in one block =1000, |
mult |
numeric: internal value, can be used to influence processing speed/ram usage |
hmax |
numeric: precut the tree at this height, default 0.3 - see ?cutreeDynamicTree |
sampNameCol |
integer: which column from the csv file contains sample names? |
collapse |
logical: reduce feature intensities to spectrum intensities? |
usePheno |
logical: transfer phenotype data from XCMS object to SpecAbund dataset? |
mspout |
logical: write msp formatted spectra to file? |
ExpDes |
either an R object created by R ExpDes object: data used for record keeping and labelling msp spectral output |
normalize |
character: either "none", "TIC", "quantile", or "batch.qc" normalization of feature intensities. see batch.qc overview in details. |
qc.inj.range |
integer: how many injections around each injection are to be scanned for presence of QC samples when using batch.qc normalization? A good rule of thumb is between 1 and 3 times the typical injection span between QC injections. i.e. if you inject QC ever 7 samples, set this to between 7 and 21. smaller values provide more local precision but make normalization sensitive to individual poor outliers (though these are first removed using the boxplot function outlier detection), while wider values provide less local precision in normalization but better stability to individual peak areas. |
order |
integer vector with length equal to number of injections in xset or csv file |
batch |
integer vector with length equal to number of injections in xset or csv file |
qc |
logical vector with length equal to number of injections in xset or csv file. |
minModuleSize |
integer: how many features must be part of a cluster to be returned? default = 2 |
linkage |
character: heirarchical clustering linkage method - see ?hclust |
mzdec |
integer: number of decimal places used in printing m/z values |
cor.method |
character: which correlational method used to calculate 'r' - see ?cor |
rt.only.low.n |
logical: default = TRUE At low injection numbers, correlational relationships of peak intensities may be unreliable. by defualt ramclustR will simply ignore the correlational r value and cluster on retention time alone. if you wish to use correlation with at n < 5, set this value to FALSE. |
replace.zeros |
logical: TRUE by default. NA, NaN, and Inf values are replaced with zero, and zero values are sometimes returned from peak peaking. When TRUE, zero values will be replaced with a small amount of noise, with noise level set based on the detected signal intensities for that feature. |
Main clustering function output - see citation for algorithm description or vignette('RAMClustR') for a walk through. batch.qc. normalization requires input of three vectors (1) batch (2) order (3) qc. This is a feature centric normalization approach which adjusts signal intensities first by comparing batch median intensity of each feature (one feature at a time) QC signal intensity to full dataset median to correct for systematic batch effects and then secondly to apply a local QC median vs global median sample correction to correct for run order effects.
$featclus: integer vector of cluster membership for each feature
$frt: feature retention time, in whatever units were fed in (xcms uses seconds, by default)
$fmz: feature retention time, reported in number of decimal points selected in ramclustR function
$xcmsOrd: the original XCMS (or csv) feature order for cross referencing, if need be
$clrt: cluster retention time
$clrtsd: retention time standard deviation of all the features that comprise that cluster
$nfeat: number of features in the cluster
$nsing: number of 'singletons' - that is the number of features which clustered with no other feature
$ExpDes: the experimental design object used when running ramclustR. List of two dataframes.
$cmpd: compound name. C#### are assigned in order of output by dynamicTreeCut. Compound with the most features is classified as C0001...
$ann: annotation. By default, annotation names are identical to 'cmpd' names. This slot is a placeholder for when annotations are provided
$MSdata: the MSdataset provided by either xcms or csv input
$MSMSdata: the (optional) MSe/idMSMS dataset provided be either xcms or csv input
$SpecAbund: the cluster intensities after collapsing features to clusters
$SpecAbundAve: the cluster intensities after averaging all samples with identical sample names
- 'spectra' directory is created in the working directory. In this directory a .msp is (optionally) created, which contains the spectra for all compounds in the dataset following clustering. if MSe/idMSMS data are provided, they are listed width he same compound name as the MS spectrum, with the collision energy provided in the ExpDes object provided to distinguish low from high CE spectra.
Corey Broeckling
Broeckling CD, Afsar FA, Neumann S, Ben-Hur A, Prenni JE. RAMClust: a novel feature clustering method enables spectral-matching-based annotation for metabolomics data. Anal Chem. 2014 Jul 15;86(14):6812-7. doi: 10.1021/ac501530d. Epub 2014 Jun 26. PubMed PMID: 24927477.
Broeckling CD, Ganna A, Layer M, Brown K, Sutton B, Ingelsson E, Peers G, Prenni JE. Enabling Efficient and Confident Annotation of LC-MS Metabolomics Data through MS1 Spectrum and Time Prediction. Anal Chem. 2016 Sep 20;88(18):9226-34. doi: 10.1021/acs.analchem.6b02479. Epub 2016 Sep 8. PubMed PMID: 7560453.
## Choose input file with feature column names `mz_rt` (expected by default). ## Column with sample name is expected to be first (by default). ## These can be adjusted with the `featdelim` and `sampNameCol` parameters. wd <- getwd() filename <- system.file("extdata", "peaks.csv", package = "RAMClustR", mustWork = TRUE) print(filename) head(data.frame(read.csv(filename)), c(6L, 5L)) ## If the file contains features from MS1, assign those to the `ms` parameter. ## If the file contains features from MS2, assign those to the `idmsms` parameter. ## If you ran `xcms` for the feature detection, the assign the output to the `xcmsObj` parameter. ## In this example we use a MS1 feature table stored in a `csv` file. setwd(tempdir()) ramclustobj <- ramclustR(ms = filename, st = 5, maxt = 1, blocksize = 1000) ## Investigate the deconvoluted features in the `spectra` folder in MSP format ## or inspect the `ramclustobj` for feature retention times, annotations etc. print(ramclustobj$ann) print(ramclustobj$nfeat) print(ramclustobj$SpecAbund[, 1:6]) setwd(wd)## Choose input file with feature column names `mz_rt` (expected by default). ## Column with sample name is expected to be first (by default). ## These can be adjusted with the `featdelim` and `sampNameCol` parameters. wd <- getwd() filename <- system.file("extdata", "peaks.csv", package = "RAMClustR", mustWork = TRUE) print(filename) head(data.frame(read.csv(filename)), c(6L, 5L)) ## If the file contains features from MS1, assign those to the `ms` parameter. ## If the file contains features from MS2, assign those to the `idmsms` parameter. ## If you ran `xcms` for the feature detection, the assign the output to the `xcmsObj` parameter. ## In this example we use a MS1 feature table stored in a `csv` file. setwd(tempdir()) ramclustobj <- ramclustR(ms = filename, st = 5, maxt = 1, blocksize = 1000) ## Investigate the deconvoluted features in the `spectra` folder in MSP format ## or inspect the `ramclustobj` for feature retention times, annotations etc. print(ramclustobj$ann) print(ramclustobj$nfeat) print(ramclustobj$SpecAbund[, 1:6]) setwd(wd)
extractor for xcms objects in preparation for clustering
rc.calibrate.ri(ramclustObj = NULL, calibrant.data = "", poly.order = 3)rc.calibrate.ri(ramclustObj = NULL, calibrant.data = "", poly.order = 3)
ramclustObj |
ramclustObj containing MSdata with optional MSMSdata (MSe, DIA, idMSMS) |
calibrant.data |
character vector defining the file path/name to a csv file containing columns including 'rt', and 'ri'. Alternatively, a data.frame with those columnn names (case sensitive) |
poly.order |
integer default = 3. polynomical order used to fit rt vs ri data, and calculate ri for all feature and metabolite rt values. poly.order should be apprciably smaller than the number of calibrant points. |
This function generates a new slot in the ramclustR object for retention index. Calibration is performed using a polynomial fit of order poly.order. It is the user's responsibility to ensure that the number and span of calibrant points is sufficient to calibrate the full range of feature and compound retention times. i.e. if the last calibration point is at 1000 seconds, but the last eluting peak is at 1300 seconds, the calibration will be very poor for the late eluting compound.
ramclustR object with retention index assigned for features ($fri) and compounds ($clri).
Corey Broeckling
Broeckling CD, Afsar FA, Neumann S, Ben-Hur A, Prenni JE. RAMClust: a novel feature clustering method enables spectral-matching-based annotation for metabolomics data. Anal Chem. 2014 Jul 15;86(14):6812-7. doi: 10.1021/ac501530d. Epub 2014 Jun 26. PubMed PMID: 24927477.
used to remove compounds which are found at similar intensity in blank samples. Only applied after clustering. see also rc.feature.filter.blanks for filtering at the feature level (only done before clustering).
rc.cmpd.filter.blanks( ramclustObj = NULL, qc.tag = "QC", blank.tag = "blank", sn = 3, remove.blanks = TRUE )rc.cmpd.filter.blanks( ramclustObj = NULL, qc.tag = "QC", blank.tag = "blank", sn = 3, remove.blanks = TRUE )
ramclustObj |
ramclustObj containing SpecAbund dataframe. |
qc.tag |
character vector of length one or two. If length is two, enter search string and factor name in $phenoData slot (i.e. c("QC", "sample.type"). If length one (i.e. "QC"), will search for this string in the 'sample.names' slot by default. |
blank.tag |
see 'qc.tag' , but for blanks to use as background. |
sn |
numeric defines the ratio for 'signal'. i.e. sn = 3 indicates that signal intensity must be 3 fold higher in sample than in blanks, on average, to be retained. |
remove.blanks |
logical. TRUE by default. this removes any recognized blanks samples from the SpecAbund sets after they are used to filter contaminant compounds |
This function removes compounds which contain signal in QC samples comparable to blanks.
ramclustR object with normalized data.
Corey Broeckling
Broeckling CD, Afsar FA, Neumann S, Ben-Hur A, Prenni JE. RAMClust: a novel feature clustering method enables spectral-matching-based annotation for metabolomics data. Anal Chem. 2014 Jul 15;86(14):6812-7. doi: 10.1021/ac501530d. Epub 2014 Jun 26. PubMed PMID: 24927477.
extractor for xcms objects in preparation for clustering
rc.cmpd.filter.cv(ramclustObj = NULL, qc.tag = "QC", max.cv = 0.5)rc.cmpd.filter.cv(ramclustObj = NULL, qc.tag = "QC", max.cv = 0.5)
ramclustObj |
ramclustObj containing MSdata with optional MSMSdata (MSe, DIA, idMSMS) |
qc.tag |
character vector of length one or two. If length is two, enter search string and factor name in $phenoData slot (i.e. c("QC", "sample.type"). If length one (i.e. "QC"), will search for this string in the 'sample.names' slot by default. |
max.cv |
numeric maximum allowable cv for any feature. default = 0.3 |
This function offers normalization by total extracted ion signal. it is recommended to first run 'rc.feature.filter.blanks' to remove non-sample derived signal.
ramclustR object with total extracted ion normalized data.
Corey Broeckling
Broeckling CD, Afsar FA, Neumann S, Ben-Hur A, Prenni JE. RAMClust: a novel feature clustering method enables spectral-matching-based annotation for metabolomics data. Anal Chem. 2014 Jul 15;86(14):6812-7. doi: 10.1021/ac501530d. Epub 2014 Jun 26. PubMed PMID: 24927477.
replaces any NA (and optionally zero) values with small signal (20
rc.cmpd.replace.na( ramclustObj = NULL, replace.int = 0.1, replace.noise = 0.1, replace.zero = TRUE )rc.cmpd.replace.na( ramclustObj = NULL, replace.int = 0.1, replace.noise = 0.1, replace.zero = TRUE )
ramclustObj |
ramclustObj containing SpecAbund dataset |
replace.int |
default = 0.2. proportion of minimum feature value to replace NA (or zero) values with |
replace.noise |
default = 0.2. proportion ofreplace.int value by which noise is added via 'jitter' |
replace.zero |
logical if TRUE, any zero values are replaced with noise as if they were NA values |
noise is added by finding for each feature the minimum detected value, multiplying that value by replace.int, then adding (replace.int*replace.noise) noise. abs() is used to ensure no negative values result.
ramclustR object with NA and zero values removed.
Corey Broeckling
Broeckling CD, Afsar FA, Neumann S, Ben-Hur A, Prenni JE. RAMClust: a novel feature clustering method enables spectral-matching-based annotation for metabolomics data. Anal Chem. 2014 Jul 15;86(14):6812-7. doi: 10.1021/ac501530d. Epub 2014 Jun 26. PubMed PMID: 24927477.
turn concatenated sample names into factors
rc.expand.sample.names( ramclustObj = NULL, delim = "-", factor.names = TRUE, quiet = FALSE )rc.expand.sample.names( ramclustObj = NULL, delim = "-", factor.names = TRUE, quiet = FALSE )
ramclustObj |
ramclustObj containing MSdata with optional MSMSdata (MSe, DIA, idMSMS) |
delim |
what delimiter should be used to separate names into factors? '-' by default |
factor.names |
logical or character vector. if TRUE, user will enter names one by on in console. If character vector (i.e. c("trt", "time")) names are assigned to table |
quiet |
logical . if TRUE, user will not be prompted to enter names one by on in console. |
THis function only works on newer format ramclustObjects with a $phenoData slot.
This function will split sample names by a delimiter, and enable users to name factors
ramclustR object with normalized data.
Corey Broeckling
Broeckling CD, Afsar FA, Neumann S, Ben-Hur A, Prenni JE. RAMClust: a novel feature clustering method enables spectral-matching-based annotation for metabolomics data. Anal Chem. 2014 Jul 15;86(14):6812-7. doi: 10.1021/ac501530d. Epub 2014 Jun 26. PubMed PMID: 24927477.
Cluster annotation function: inference of 'M' - molecular weight of the compound giving rise to each spectrum - using the InterpretMSSpectrum::findMain function
rc.export.msp.rc(ramclustObj = NULL, one.file = TRUE, mzdec = 1)rc.export.msp.rc(ramclustObj = NULL, one.file = TRUE, mzdec = 1)
ramclustObj |
ramclustR object to annotate. |
one.file |
logical, should all msp spectra be written to one file? If false, each spectrum is an individual file. |
mzdec |
integer. Number of decimal points to export mass values with. |
exports files to a directory called 'spectra'. If one.file = FALSE, a new directory 'spectra/msp' is created to hold the individual msp files. if do.findman has been run, spectra are written as ms2 spectra, else as ms1.
nothing, just exports files to the working directory
Corey Broeckling
used to remove features which are found at similar intensity in blank samples
rc.feature.filter.blanks( ramclustObj = NULL, qc.tag = "QC", blank.tag = "blank", sn = 3, remove.blanks = TRUE )rc.feature.filter.blanks( ramclustObj = NULL, qc.tag = "QC", blank.tag = "blank", sn = 3, remove.blanks = TRUE )
ramclustObj |
ramclustObj containing MSdata with optional MSMSdata (MSe, DIA, idMSMS) |
qc.tag |
character vector of length one or two. If length is two, enter search string and factor name in $phenoData slot (i.e. c("QC", "sample.type"). If length one (i.e. "QC"), will search for this string in the 'sample.names' slot by default. |
blank.tag |
see 'qc.tag' , but for blanks to use as background. |
sn |
numeric defines the ratio for 'signal'. i.e. sn = 3 indicates that signal intensity must be 3 fold higher in sample than in blanks, on average, to be retained. |
remove.blanks |
logical. TRUE by default. this removes any recognized blanks samples from the MSdata and MSMSdata sets after they are used to filter contaminant features. |
This function offers normalization by run order, batch number, and QC sample signal intensity.
Each input vector should be the same length, and equal to the number of samples in the $MSdata set.
Input vector order is assumed to be the same as the sample order in the $MSdata set.
ramclustR object with normalized data.
Corey Broeckling
Broeckling CD, Afsar FA, Neumann S, Ben-Hur A, Prenni JE. RAMClust: a novel feature clustering method enables spectral-matching-based annotation for metabolomics data. Anal Chem. 2014 Jul 15;86(14):6812-7. doi: 10.1021/ac501530d. Epub 2014 Jun 26. PubMed PMID: 24927477.
extractor for xcms objects in preparation for clustering
rc.feature.filter.cv(ramclustObj = NULL, qc.tag = "QC", max.cv = 0.5)rc.feature.filter.cv(ramclustObj = NULL, qc.tag = "QC", max.cv = 0.5)
ramclustObj |
ramclustObj containing MSdata with optional MSMSdata (MSe, DIA, idMSMS) |
qc.tag |
character vector of length one or two. If length is two, enter search string and factor name in $phenoData slot (i.e. c("QC", "sample.type"). If length one (i.e. "QC"), will search for this string in the 'sample.names' slot by default. |
max.cv |
numeric maximum allowable cv for any feature. default = 0.5 |
This function offers normalization by total extracted ion signal. it is recommended to first run 'rc.feature.filter.blanks' to remove non-sample derived signal.
ramclustR object with total extracted ion normalized data.
Corey Broeckling
Broeckling CD, Afsar FA, Neumann S, Ben-Hur A, Prenni JE. RAMClust: a novel feature clustering method enables spectral-matching-based annotation for metabolomics data. Anal Chem. 2014 Jul 15;86(14):6812-7. doi: 10.1021/ac501530d. Epub 2014 Jun 26. PubMed PMID: 24927477.
normalize data using batch.qc
rc.feature.normalize.batch.qc( order = NULL, batch = NULL, qc = NULL, ramclustObj = NULL, qc.inj.range = 20 )rc.feature.normalize.batch.qc( order = NULL, batch = NULL, qc = NULL, ramclustObj = NULL, qc.inj.range = 20 )
order |
integer vector with length equal to number of injections in xset or csv file or dataframe |
batch |
integer vector with length equal to number of injections in xset or csv file or dataframe |
qc |
logical vector with length equal to number of injections in xset or csv file or dataframe |
ramclustObj |
ramclustObj containing MSdata with optional MSMSdata (MSe, DIA, idMSMS) |
qc.inj.range |
integer: how many injections around each injection are to be scanned for presence of QC samples when using batch.qc normalization? A good rule of thumb is between 1 and 3 times the typical injection span between QC injections. i.e. if you inject QC ever 7 samples, set this to between 7 and 21. smaller values provide more local precision but make normalization sensitive to individual poor outliers (though these are first removed using the boxplot function outlier detection), while wider values provide less local precision in normalization but better stability to individual peak areas. |
ramclustR object with normalized data.
extractor for xcms objects in preparation for clustering
rc.feature.normalize.qc( ramclustObj = NULL, order = NULL, batch = NULL, qc.tag = NULL, output.plot = FALSE, p.cut = 0.05, rsq.cut = 0.1, p.adjust = "none" )rc.feature.normalize.qc( ramclustObj = NULL, order = NULL, batch = NULL, qc.tag = NULL, output.plot = FALSE, p.cut = 0.05, rsq.cut = 0.1, p.adjust = "none" )
ramclustObj |
ramclustObj containing MSdata with optional MSMSdata (MSe, DIA, idMSMS) |
order |
integer vector with length equal to number of injections in xset or csv file |
batch |
integer vector with length equal to number of injections in xset or csv file |
qc.tag |
character vector of length one or two. If length is two, enter search string and factor name in $phenoData slot (i.e. c("QC", "sample.type"). If length one (i.e. "QC"), will search for this string in the 'sample.names' slot by default. |
output.plot |
logical: if TRUE (default), plots are output to PDF. |
p.cut |
numeric when run order correction is applied, only features showing a run order vs signal with a linear p-value (after FDR correction) < p.cut will be adjusted. also requires r-squared < rsq.cut. |
rsq.cut |
numeric when run order correction is applied, only features showing a run order vs signal with a linear r-squared > rsq.cut will be adjusted. also requires p values < p.cut. |
p.adjust |
which p-value adjustment should be used? default = "none", see ?p.adjust |
This function offers normalization by run order, batch number, and QC sample signal intensity.
Each input vector should be the same length, and equal to the number of samples in the $MSdata set.
Input vector order is assumed to be the same as the sample order in the $MSdata set.
ramclustR object with normalized data.
Corey Broeckling
Broeckling CD, Afsar FA, Neumann S, Ben-Hur A, Prenni JE. RAMClust: a novel feature clustering method enables spectral-matching-based annotation for metabolomics data. Anal Chem. 2014 Jul 15;86(14):6812-7. doi: 10.1021/ac501530d. Epub 2014 Jun 26. PubMed PMID: 24927477.
normalize data using quantile
rc.feature.normalize.quantile(ramclustObj = NULL)rc.feature.normalize.quantile(ramclustObj = NULL)
ramclustObj |
ramclustObj containing MSdata with optional MSMSdata (MSe, DIA, idMSMS) |
ramclustR object with normalized data.
extractor for xcms objects in preparation for clustering
rc.feature.normalize.tic(ramclustObj = NULL)rc.feature.normalize.tic(ramclustObj = NULL)
ramclustObj |
ramclustObj containing MSdata with optional MSMSdata (MSe, DIA, idMSMS) |
This function offers normalization by total extracted ion signal. it is recommended to first run 'rc.feature.filter.blanks' to remove non-sample derived signal.
ramclustR object with total extracted ion normalized data.
Corey Broeckling
Broeckling CD, Afsar FA, Neumann S, Ben-Hur A, Prenni JE. RAMClust: a novel feature clustering method enables spectral-matching-based annotation for metabolomics data. Anal Chem. 2014 Jul 15;86(14):6812-7. doi: 10.1021/ac501530d. Epub 2014 Jun 26. PubMed PMID: 24927477.
replaces any NA (and optionally zero) values with small signal (20
rc.feature.replace.na( ramclustObj = NULL, replace.int = 0.1, replace.noise = 0.1, replace.zero = TRUE, which.data = c("MSdata", "MSMSdata") )rc.feature.replace.na( ramclustObj = NULL, replace.int = 0.1, replace.noise = 0.1, replace.zero = TRUE, which.data = c("MSdata", "MSMSdata") )
ramclustObj |
ramclustObj containing MSdata with optional MSMSdata (MSe, DIA, idMSMS) |
replace.int |
default = 0.1. proportion of minimum feature value to replace NA (or zero) values with |
replace.noise |
default = 0.1. proportion ofreplace.int value by which noise is added via 'jitter' |
replace.zero |
logical if TRUE, any zero values are replaced with noise as if they were NA values |
which.data |
name of dataset |
noise is added by finding for each feature the minimum detected value, multiplying that value by replace.int, then adding (replace.int*replace.noise) noise. abs() is used to ensure no negative values result.
ramclustR object with NA and zero values removed.
Corey Broeckling
Broeckling CD, Afsar FA, Neumann S, Ben-Hur A, Prenni JE. RAMClust: a novel feature clustering method enables spectral-matching-based annotation for metabolomics data. Anal Chem. 2014 Jul 15;86(14):6812-7. doi: 10.1021/ac501530d. Epub 2014 Jun 26. PubMed PMID: 24927477.
extractor for csv objects in preparation for normalization and clustering
rc.get.csv.data( csv = NULL, phenoData = NULL, idmsms = NULL, ExpDes = NULL, sampNameCol = 1, st = NULL, timepos = 2, featdelim = "_", ensure.no.na = TRUE )rc.get.csv.data( csv = NULL, phenoData = NULL, idmsms = NULL, ExpDes = NULL, sampNameCol = 1, st = NULL, timepos = 2, featdelim = "_", ensure.no.na = TRUE )
csv |
filepath: csv input. Features as columns, rows as samples. Column header mz_rt |
phenoData |
character: character string in 'taglocation' to designate files as either MS / DIA(MSe, MSall, AIF, etc) e.g. "01.mzML" |
idmsms |
filepath: optional idMSMS / MSe csv data. same dim and names as ms required |
ExpDes |
either an R object created by R ExpDes object: data used for record keeping and labelling msp spectral output |
sampNameCol |
integer: which column from the csv file contains sample names? |
st |
numeric: sigma t - time similarity decay value |
timepos |
integer: which position in delimited column header represents the retention time |
featdelim |
character: how feature mz and rt are delimited in csv import column header e.g. ="-" |
ensure.no.na |
logical: if TRUE, any 'NA' values in msint and/or msmsint are replaced with numerical values based on 10 percent of feature min plus noise. Used to ensure that spectra are not written with NA values. |
This function creates a ramclustObj which will be used as input for clustering.
an empty ramclustR object. this object is formatted as an hclust object with additional slots for holding feature and compound data. details on these found below.
$frt: feature retention time, in whatever units were fed in
$fmz: feature retention time, reported in number of decimal points selected in ramclustR function
$ExpDes: the experimental design object used when running ramclustR. List of two dataframes.
$MSdata: the MSdataset provided by either xcms or csv input
$MSMSdata: the (optional) DIA(MSe, MSall, AIF etc) dataset
$xcmsOrd: original xcms order of features, for back-referencing when necessary
$msint: weighted.mean intensity of feature in ms level data
$msmsint:weighted.mean intensity of feature in msms level data
Corey Broeckling
Broeckling CD, Afsar FA, Neumann S, Ben-Hur A, Prenni JE. RAMClust: a novel feature clustering method enables spectral-matching-based annotation for metabolomics data. Anal Chem. 2014 Jul 15;86(14):6812-7. doi: 10.1021/ac501530d. Epub 2014 Jun 26. PubMed PMID: 24927477.
Broeckling CD, Ganna A, Layer M, Brown K, Sutton B, Ingelsson E, Peers G, Prenni JE. Enabling Efficient and Confident Annotation of LC-MS Metabolomics Data through MS1 Spectrum and Time Prediction. Anal Chem. 2016 Sep 20;88(18):9226-34. doi: 10.1021/acs.analchem.6b02479. Epub 2016 Sep 8. PubMed PMID: 7560453.
## Choose csv input file. Features as columns, rows as samples ## Choose csv input file phenoData filename <- system.file("extdata", "peaks.csv", package = "RAMClustR", mustWork = TRUE) phenoData <- system.file("extdata", "phenoData.csv", package = "RAMClustR", mustWork = TRUE) ramclustobj <- rc.get.csv.data(csv = filename, phenoData = phenoData, st = 5)## Choose csv input file. Features as columns, rows as samples ## Choose csv input file phenoData filename <- system.file("extdata", "peaks.csv", package = "RAMClustR", mustWork = TRUE) phenoData <- system.file("extdata", "phenoData.csv", package = "RAMClustR", mustWork = TRUE) ramclustobj <- rc.get.csv.data(csv = filename, phenoData = phenoData, st = 5)
extractor for dataframe input in preparation for normalization and clustering
rc.get.df.data( ms1_featureDefinitions = NULL, ms1_featureValues = NULL, ms2_featureDefinitions = NULL, ms2_featureValues = NULL, phenoData = NULL, ExpDes = NULL, featureNamesColumnIndex = 1, st = NULL, ensure.no.na = TRUE )rc.get.df.data( ms1_featureDefinitions = NULL, ms1_featureValues = NULL, ms2_featureDefinitions = NULL, ms2_featureValues = NULL, phenoData = NULL, ExpDes = NULL, featureNamesColumnIndex = 1, st = NULL, ensure.no.na = TRUE )
ms1_featureDefinitions |
dataframe with metadata with columns: mz, rt, feature names containing MS data |
ms1_featureValues |
dataframe with rownames = sample names, colnames = feature names containing MS data |
ms2_featureDefinitions |
dataframe with metadata with columns: mz, rt, feature names containing MSMS data |
ms2_featureValues |
dataframe with rownames = sample names, colnames = feature names containing MSMS data |
phenoData |
dataframe containing phenoData |
ExpDes |
either an R object created by R ExpDes object: data used for record keeping and labelling msp spectral output |
featureNamesColumnIndex |
integer: which column in 'ms1_featureDefinitions' contains feature names? |
st |
numeric: sigma t - time similarity decay value |
ensure.no.na |
logical: if TRUE, any 'NA' values in msint and/or msmsint are replaced with numerical values based on 10 percent of feature min plus noise. Used to ensure that spectra are not written with NA values. |
This function creates a ramclustObj which will be used as input for clustering.
an empty ramclustR object. this object is formatted as an hclust object with additional slots for holding feature and compound data. details on these found below.
$frt: feature retention time, in whatever units were fed in
$fmz: feature retention time, reported in number of decimal points selected in ramclustR function
$ExpDes: the experimental design object used when running ramclustR. List of two dataframes.
$MSdata: the MSdataset provided by either xcms or csv input
$MSMSdata: the (optional) DIA(MSe, MSall, AIF etc) dataset
$xcmsOrd: original xcms order of features, for back-referencing when necessary
$msint: weighted.mean intensity of feature in ms level data
$msmsint:weighted.mean intensity of feature in msms level data
Zargham Ahmad, Helge Hecht, Corey Broeckling
Broeckling CD, Afsar FA, Neumann S, Ben-Hur A, Prenni JE. RAMClust: a novel feature clustering method enables spectral-matching-based annotation for metabolomics data. Anal Chem. 2014 Jul 15;86(14):6812-7. doi: 10.1021/ac501530d. Epub 2014 Jun 26. PubMed PMID: 24927477.
Broeckling CD, Ganna A, Layer M, Brown K, Sutton B, Ingelsson E, Peers G, Prenni JE. Enabling Efficient and Confident Annotation of LC-MS Metabolomics Data through MS1 Spectrum and Time Prediction. Anal Chem. 2016 Sep 20;88(18):9226-34. doi: 10.1021/acs.analchem.6b02479. Epub 2016 Sep 8. PubMed PMID: 7560453.
## Choose dataframe with metadata with columns: mz, rt, feature names containing MS data ## Choose dataframe with rownames = sample names, colnames = feature names containing MS data ## Choose dataframe containing phenoData df1 <- readRDS(system.file("extdata", "featDefinition.rds", package = "RAMClustR", mustWork = TRUE)) df2 <- readRDS(system.file("extdata", "featValues.rds", package = "RAMClustR", mustWork = TRUE)) df3 <- readRDS(system.file("extdata", "phenoData_df.rds", package = "RAMClustR", mustWork = TRUE)) ramclustr <- rc.get.df.data(ms1_featureDefinitions=df1, ms1_featureValues=df2, phenoData=df3, st=5)## Choose dataframe with metadata with columns: mz, rt, feature names containing MS data ## Choose dataframe with rownames = sample names, colnames = feature names containing MS data ## Choose dataframe containing phenoData df1 <- readRDS(system.file("extdata", "featDefinition.rds", package = "RAMClustR", mustWork = TRUE)) df2 <- readRDS(system.file("extdata", "featValues.rds", package = "RAMClustR", mustWork = TRUE)) df3 <- readRDS(system.file("extdata", "phenoData_df.rds", package = "RAMClustR", mustWork = TRUE)) ramclustr <- rc.get.df.data(ms1_featureDefinitions=df1, ms1_featureValues=df2, phenoData=df3, st=5)
extractor for xcms objects in preparation for normalization and clustering
rc.get.xcms.data( xcmsObj = NULL, taglocation = "filepaths", MStag = NULL, MSMStag = NULL, ExpDes = NULL, mzdec = 3, ensure.no.na = TRUE )rc.get.xcms.data( xcmsObj = NULL, taglocation = "filepaths", MStag = NULL, MSMStag = NULL, ExpDes = NULL, mzdec = 3, ensure.no.na = TRUE )
xcmsObj |
xcmsObject: containing grouped feature data for clustering by ramclustR |
taglocation |
character: "filepaths" by default, "phenoData[,1]" is another option. refers to xcms slot |
MStag |
character: character string in 'taglocation' to designate files as either MS / DIA(MSe, MSall, AIF, etc) e.g. "01.mzML" |
MSMStag |
character: character string in 'taglocation' to designate files as either MS / DIA(MSe, MSall, AIF, etc) e.g. "02.mzML" |
ExpDes |
either an R object created by R ExpDes object: data used for record keeping and labelling msp spectral output |
mzdec |
integer: number of decimal places for storing m/z values |
ensure.no.na |
logical: if TRUE, any 'NA' values in msint and/or msmsint are replaced with numerical values based on 10 percent of feature min plus noise. Used to ensure that spectra are not written with NA values. |
This function creates a ramclustObj which will be used as input for clustering.
an empty ramclustR object. this object is formatted as an hclust object with additional slots for holding feature and compound data. details on these found below.
$frt: feature retention time, in whatever units were fed in (xcms uses seconds, by default)
$fmz: feature retention time, reported in number of decimal points selected in ramclustR function
$ExpDes: the experimental design object used when running ramclustR. List of two dataframes.
$MSdata: the MSdataset provided by either xcms or csv input
$MSMSdata: the (optional) DIA(MSe, MSall, AIF etc) dataset provided be either xcms or csv input
$xcmsOrd: original xcms order of features, for back-referencing when necessary
$msint: weighted.mean intensity of feature in ms level data
$msmsint:weighted.mean intensity of feature in msms level data
Corey Broeckling
Broeckling CD, Afsar FA, Neumann S, Ben-Hur A, Prenni JE. RAMClust: a novel feature clustering method enables spectral-matching-based annotation for metabolomics data. Anal Chem. 2014 Jul 15;86(14):6812-7. doi: 10.1021/ac501530d. Epub 2014 Jun 26. PubMed PMID: 24927477.
Broeckling CD, Ganna A, Layer M, Brown K, Sutton B, Ingelsson E, Peers G, Prenni JE. Enabling Efficient and Confident Annotation of LC-MS Metabolomics Data through MS1 Spectrum and Time Prediction. Anal Chem. 2016 Sep 20;88(18):9226-34. doi: 10.1021/acs.analchem.6b02479. Epub 2016 Sep 8. PubMed PMID: 7560453.
Cluster refinement - scanning instruments (quadrupole, as in GC-MS) can display cluster splitting, possibily due to slight differences in measured peak retentiont time as a function of mass due to scan dynamics. this function enables a second pass clustering designed to merge two clusters if the second cluster is within a small retention time window and shows a sufficiently strong correlation.
rc.merge.split.clusters( ramclustObj = NULL, merge.threshold = 0.7, cor.method = "spearman", rt.sd.factor = 3 )rc.merge.split.clusters( ramclustObj = NULL, merge.threshold = 0.7, cor.method = "spearman", rt.sd.factor = 3 )
ramclustObj |
ramclustR object to annotate. |
merge.threshold |
numeric. value between -1 and 1 indicating the correlational r threshold above which two clusters will be merged |
cor.method |
character. default = 'spearman'. correlational method to use for calculating r. see documentation on R base cor() function for available options |
rt.sd.factor |
numeric. default = 3. clusters within rt.sd.factor * ramclustObj$rtsd (cluster retention time standard deviation) are considered for merging. |
exports files to a directory called 'spectra'. If one.file = FALSE, a new directory 'spectra/msp' is created to hold the individual msp files. if do.findman has been run, spectra are written as ms2 spectra, else as ms1.
new ramclustR object, with (generally) fewer clusters than the input ramclustR object.
Corey Broeckling
summarize quality control for clustering and for quality control sample variation based on compound ($SpecAbund) and feature ($MSdata and $MSMSdata, if present)
rc.qc( ramclustObj = NULL, qc.tag = "QC", remove.qc = FALSE, npc = 4, scale = "pareto", outfile.basename = "ramclustQC", view.hist = TRUE )rc.qc( ramclustObj = NULL, qc.tag = "QC", remove.qc = FALSE, npc = 4, scale = "pareto", outfile.basename = "ramclustQC", view.hist = TRUE )
ramclustObj |
ramclustR object to analyze |
qc.tag |
qc.tag character vector of length one or two. If length is two, enter search string and factor name in $phenoData slot (i.e. c("QC", "sample.type"). If length one (i.e. "QC"), will search for this string in the 'sample.names' slot by default. |
remove.qc |
logical - if TRUE (default) QC injections will be removed from the returned ramclustObj (applies to $MSdata, $MSMSdata, $SpecAbund, $phenoData, as appropriate). If FALSE, QC samples remain. |
npc |
number of Principle components to calcuate and plot |
scale |
"pareto" by default: PCA scaling method used |
outfile.basename |
base name of output files. Extensions added internally. default = "ramclustQC" |
view.hist |
logical. should histograms be plotted? |
plots a ramclustR summary plot. first page represents the correlation of each cluster to all other clusters, sorted by retention time. large blocks of yellow along the diaganol indicate either poor clustering or a group of coregulated metabolites with similar retention time. It is an imperfect diagnostic, particularly with lipids on reverse phase LC or sugars on HILIC LC systems. Page 2: histogram of r values from page 1 - only r values one position from the diagonal are used. Pages 3:5 - PCA results, with QC samples colored red. relative standard deviation calculated as sd(QC PC scores) / sd(all PC scores). Page 6: histogram of CV values for each compound int he dataset, QC samples only.
new RC object. Saves output summary plots to pdf and .csv summary tables to new 'QC' directory. If remove.qc = TRUE, moves QC samples to new $QC slot from original position.
Corey Broeckling
Broeckling CD, Afsar FA, Neumann S, Ben-Hur A, Prenni JE. RAMClust: a novel feature clustering method enables spectral-matching-based annotation for metabolomics data. Anal Chem. 2014 Jul 15;86(14):6812-7. doi: 10.1021/ac501530d. Epub 2014 Jun 26. PubMed PMID: 24927477.
Broeckling CD, Ganna A, Layer M, Brown K, Sutton B, Ingelsson E, Peers G, Prenni JE. Enabling Efficient and Confident Annotation of LC-MS Metabolomics Data through MS1 Spectrum and Time Prediction. Anal Chem. 2016 Sep 20;88(18):9226-34. doi: 10.1021/acs.analchem.6b02479. Epub 2016 Sep 8. PubMed PMID: 7560453.
Main clustering function for grouping features based on their analytical behavior.
rc.ramclustr( ramclustObj = NULL, st = NULL, sr = NULL, maxt = NULL, deepSplit = FALSE, blocksize = 2000, mult = 5, hmax = NULL, collapse = TRUE, minModuleSize = 2, linkage = "average", cor.method = "pearson", rt.only.low.n = TRUE )rc.ramclustr( ramclustObj = NULL, st = NULL, sr = NULL, maxt = NULL, deepSplit = FALSE, blocksize = 2000, mult = 5, hmax = NULL, collapse = TRUE, minModuleSize = 2, linkage = "average", cor.method = "pearson", rt.only.low.n = TRUE )
ramclustObj |
ramclustR object: containing ungrouped features. constructed by rc.get.xcms.data, for example |
st |
numeric: sigma t - time similarity decay value |
sr |
numeric: sigma r - correlational similarity decay value |
maxt |
numeric: maximum time difference to calculate retention similarity for - all values beyond this are assigned similarity of zero |
deepSplit |
logical: controls how agressively the HCA tree is cut - see ?cutreeDynamicTree |
blocksize |
integer: number of features (scans?) processed in one block =1000, |
mult |
numeric: internal value, can be used to influence processing speed/ram usage |
hmax |
numeric: precut the tree at this height, default 0.3 - see ?cutreeDynamicTree |
collapse |
logical: if true (default), feature quantitative values are collapsed into spectra quantitative values. |
minModuleSize |
integer: how many features must be part of a cluster to be returned? default = 2 |
linkage |
character: heirarchical clustering linkage method - see ?hclust |
cor.method |
character: which correlational method used to calculate 'r' - see ?cor |
rt.only.low.n |
logical: default = TRUE At low injection numbers, correlational relationships of peak intensities may be unreliable. by defualt ramclustR will simply ignore the correlational r value and cluster on retention time alone. if you wish to use correlation with at n < 5, set this value to FALSE. |
Main clustering function output - see citation for algorithm description or vignette('RAMClustR') for a walk through. batch.qc. normalization requires input of three vectors (1) batch (2) order (3) qc. This is a feature centric normalization approach which adjusts signal intensities first by comparing batch median intensity of each feature (one feature at a time) QC signal intensity to full dataset median to correct for systematic batch effects and then secondly to apply a local QC median vs global median sample correction to correct for run order effects.
$featclus: integer vector of cluster membership for each feature
$clrt: cluster retention time
$clrtsd: retention time standard deviation of all the features that comprise that cluster
$nfeat: number of features in the cluster
$nsing: number of 'singletons' - that is the number of features which clustered with no other feature
$cmpd: compound name. C#### are assigned in order of output by dynamicTreeCut. Compound with the most features is classified as C0001...
$ann: annotation. By default, annotation names are identical to 'cmpd' names. This slot is a placeholder for when annotations are provided
$SpecAbund: the cluster intensities after collapsing features to clusters
$SpecAbundAve: the cluster intensities after averaging all samples with identical sample names
Corey Broeckling
Broeckling CD, Afsar FA, Neumann S, Ben-Hur A, Prenni JE. RAMClust: a novel feature clustering method enables spectral-matching-based annotation for metabolomics data. Anal Chem. 2014 Jul 15;86(14):6812-7. doi: 10.1021/ac501530d. Epub 2014 Jun 26. PubMed PMID: 24927477.
Broeckling CD, Ganna A, Layer M, Brown K, Sutton B, Ingelsson E, Peers G, Prenni JE. Enabling Efficient and Confident Annotation of LC-MS Metabolomics Data through MS1 Spectrum and Time Prediction. Anal Chem. 2016 Sep 20;88(18):9226-34. doi: 10.1021/acs.analchem.6b02479. Epub 2016 Sep 8. PubMed PMID: 7560453.
summarize quality control for clustering and for quality control sample variation based on compound ($SpecAbund) and feature ($MSdata and $MSMSdata, if present)
rc.remove.qc(ramclustObj = NULL, qc.tag = "QC")rc.remove.qc(ramclustObj = NULL, qc.tag = "QC")
ramclustObj |
ramclustR object to analyze |
qc.tag |
qc.tag character vector of length one or two. If length is two, enter search string and factor name in $phenoData slot (i.e. c("QC", "sample.type"). If length one (i.e. "QC"), will search for this string in the 'sample.names' slot by default. |
simply moves QC samples out of the way for downstream processing. moved to a $qc slot.
new RC object. moves QC samples to new $qc slot from original position.
Corey Broeckling
Broeckling CD, Afsar FA, Neumann S, Ben-Hur A, Prenni JE. RAMClust: a novel feature clustering method enables spectral-matching-based annotation for metabolomics data. Anal Chem. 2014 Jul 15;86(14):6812-7. doi: 10.1021/ac501530d. Epub 2014 Jun 26. PubMed PMID: 24927477.
Broeckling CD, Ganna A, Layer M, Brown K, Sutton B, Ingelsson E, Peers G, Prenni JE. Enabling Efficient and Confident Annotation of LC-MS Metabolomics Data through MS1 Spectrum and Time Prediction. Anal Chem. 2016 Sep 20;88(18):9226-34. doi: 10.1021/acs.analchem.6b02479. Epub 2016 Sep 8. PubMed PMID: 7560453.
summarize quality control for clustering and for quality control sample variation based on compound ($SpecAbund) and feature ($MSdata and $MSMSdata, if present)
rc.restore.qc.samples(ramclustObj = NULL)rc.restore.qc.samples(ramclustObj = NULL)
ramclustObj |
ramclustR object to analyze |
moves all of $phenoData, $MSdata, $MSMSdata, $SpecAbund back to original positions from $qc slot
RC object
Corey Broeckling
Broeckling CD, Afsar FA, Neumann S, Ben-Hur A, Prenni JE. RAMClust: a novel feature clustering method enables spectral-matching-based annotation for metabolomics data. Anal Chem. 2014 Jul 15;86(14):6812-7. doi: 10.1021/ac501530d. Epub 2014 Jun 26. PubMed PMID: 24927477.
Broeckling CD, Ganna A, Layer M, Brown K, Sutton B, Ingelsson E, Peers G, Prenni JE. Enabling Efficient and Confident Annotation of LC-MS Metabolomics Data through MS1 Spectrum and Time Prediction. Anal Chem. 2016 Sep 20;88(18):9226-34. doi: 10.1021/acs.analchem.6b02479. Epub 2016 Sep 8. PubMed PMID: 7560453.
filter RC object and summarize quality control sample variation
RCQC( ramclustObj = NULL, qctag = "QC", npc = 4, scale = "pareto", which.data = "SpecAbund", outfile = "ramclustQC.pdf" )RCQC( ramclustObj = NULL, qctag = "QC", npc = 4, scale = "pareto", which.data = "SpecAbund", outfile = "ramclustQC.pdf" )
ramclustObj |
ramclustR object to analyze |
qctag |
"QC" by default - rowname tag to identify QC samples |
npc |
number of Principle components to calcuate and plot |
scale |
"pareto" by default: PCA scaling method used |
which.data |
which dataset to use. "SpecAbund" by default |
outfile |
name of output pdf file. |
plots a ramclustR summary plot. first page represents the correlation of each cluster to all other clusters, sorted by retention time. large blocks of yellow along the diaganol indicate either poor clustering or a group of coregulated metabolites with similar retention time. It is an imperfect diagnostic, particularly with lipids on reverse phase LC or sugars on HILIC LC systems. Page 2: histogram of r values from page 1 - only r values one position from the diagonal are used. Pages 3:5 - PCA results, with QC samples colored red. relative standard deviation calculated as sd(QC PC scores) / sd(all PC scores). Page 6: histogram of CV values for each compound int he dataset, QC samples only.
new RC object, with QC samples moved to new slot. prints output summary plots to pdf.
Corey Broeckling
Broeckling CD, Afsar FA, Neumann S, Ben-Hur A, Prenni JE. RAMClust: a novel feature clustering method enables spectral-matching-based annotation for metabolomics data. Anal Chem. 2014 Jul 15;86(14):6812-7. doi: 10.1021/ac501530d. Epub 2014 Jun 26. PubMed PMID: 24927477.
Broeckling CD, Ganna A, Layer M, Brown K, Sutton B, Ingelsson E, Peers G, Prenni JE. Enabling Efficient and Confident Annotation of LC-MS Metabolomics Data through MS1 Spectrum and Time Prediction. Anal Chem. 2016 Sep 20;88(18):9226-34. doi: 10.1021/acs.analchem.6b02479. Epub 2016 Sep 8. PubMed PMID: 7560453.
remove blanks
remove_blanks(ramclustObj, blank)remove_blanks(ramclustObj, blank)
ramclustObj |
ramclustObj containing MSdata with optional MSMSdata (MSe, DIA, idMSMS) |
blank |
blank samples found by define_samples |
ramclustObj object with blanks removed
add rc.feature.replace.na params in ramclustObj
replace_na(data, replace.int, replace.zero, replace.noise)replace_na(data, replace.int, replace.zero, replace.noise)
data |
selected data frame to use |
replace.int |
default = 0.1. proportion of minimum feature value to replace NA (or zero) values with |
replace.zero |
logical if TRUE, any zero values are replaced with noise as if they were NA values |
replace.noise |
default = 0.1. proportion ofreplace.int value by which noise is added via 'jitter' |
selected ramclustR data frame with NA and zero values removed.
number of features replaced
write csv template called "ExpDes.csv" to your working directory. you will fill this in manually, ensuring that when you save you retain csv format. ramclustR will then read this file in and and format appropriately.
write_csv(data)write_csv(data)
data |
csv template to write |
read ExpDes.csv file
Export GC-MS EI spectra for spectral searching in MSFinder
write.gcei.mat(ramclustObj = NULL)write.gcei.mat(ramclustObj = NULL)
ramclustObj |
ramclustR object to annotate. |
exports files to a directory called 'spectra'. a new directory 'spectra/mat' is created to hold the individual mat files.
nothing, just exports files to the working directory
Corey Broeckling
write RAMClustR processing methods and citations to text file
write.methods(ramclustObj = NULL, filename = NULL)write.methods(ramclustObj = NULL, filename = NULL)
ramclustObj |
R object - the ramclustR object which was used to write the .mat or .msp files |
filename |
define filename/path to write. uses 'ramclustr_methods.txt' and the working directory by default. |
this function exports a file called ramclustr_methods.txt which contains the processing history, parameters used, and relevant citations.
an annotated ramclustR object
nothing - new file written to working director
Corey Broeckling
Broeckling CD, Afsar FA, Neumann S, Ben-Hur A, Prenni JE. RAMClust: a novel feature clustering method enables spectral-matching-based annotation for metabolomics data. Anal Chem. 2014 Jul 15;86(14):6812-7. doi: 10.1021/ac501530d. Epub 2014 Jun 26. PubMed PMID: 24927477.
Cluster annotation function: inference of 'M' - molecular weight of the compound giving rise to each spectrum - using the InterpretMSSpectrum::findMain function
write.msp(ramclustObj = NULL, one.file = FALSE)write.msp(ramclustObj = NULL, one.file = FALSE)
ramclustObj |
ramclustR object to annotate. |
one.file |
logical, should all msp spectra be written to one file? If false, each spectrum is an individual file. |
exports files to a directory called 'spectra'. If one.file = FALSE, a new directory 'spectra/msp' is created to hold the individual msp files. if do.findman has been run, spectra are written as ms2 spectra, else as ms1.
nothing, just exports files to the working directory
Corey Broeckling