make.par.file {FTICRMS} | R Documentation |
Creates a file of parameters that can be read by the functions in the FTICRMS
package
make.par.file(covariates, form, root.dir = ".", par.file = "parameters.RData", ...)
covariates |
data frame with rownames given by raw data files |
form |
object of class “formula ” to be used for testing using covariates |
root.dir |
string containing location for file |
par.file |
string containing name of file |
... |
parameters whose default values are to be overwritten (see below) |
Creates a file called parameters.RData
in directory root.dir
which contains
values for all of the parameters used in the programs in the FTICRMS
package. The
possible parameters, their default values, and their descriptions are as follows:
add.norm = TRUE | logical; whether to normalize additively or multiplicatively on the log scale |
add.par = 10 | additive parameter for "shiftedlog" or "glog" options for trans.method |
align.method = "spline" | alignment algorithm for peaks |
base.dir = paste(root.dir, "/Baseline_Corrected", sep="") | directory for baseline-corrected files |
binsize = 64 | size of bins used in estimating noise variance in baseline calculation |
cor.thresh = 0.8 | threshhold correlation for declaring isotopes |
FDR = 0.1 | False Discovery Rate in Benjamini-Hochberg test |
frac.changed = 0.001 | convergence criterion in baseline calculation |
isotope.dist = 7 | maximum distance for declaring isotopes |
masses = NULL | specific masses to test |
max.iter = 30 | convergence criterion in baseline calculation |
neg.pen = sqrt(pi/2) | negativity penalty in baseline calculation |
normalization = "common" | type of normalization to use on spectra before statistical anlaysis |
num.pts = 5 | number of points needed for peak fitting |
numsds = 4 | number of standard deviations above the mean to be declared non-noise |
oneside.min = 1 | minimum number of points on each side of local maximum for peak fitting |
overwrite = FALSE | whether to replace exisiting files with new ones |
peak.dir = paste(root.dir, "/All_Peaks", sep="") | directory for peak location files |
peak.method = "parabola" | method for locating peaks |
pre.align = 0 | shifts to apply before running run.strong.peaks |
R2.thresh = 0.98 | R^2 value needed for peak fitting |
raw.dir = paste(root.dir, "/Raw_Data", sep="") | directory for raw data files |
repl.method = max | how to deal with replicates |
res.dir = paste(root.dir, "/Results", sep="") | directory for result file |
root.dir = "." | directory for parameters file and raw data |
lrg.dir = paste(root.dir, "/Large_Peaks", sep="") | directory for significant peaks file |
lrg.only = TRUE | whether to consider only peaks that have at least one peak “significant”; i.e., identified by run.lrg.peaks |
sm.fac = 10^15 | smoothing factor for baseline calculation |
trans.method = "shiftedlog" | data transformation method |
use.t.test = TRUE | whether to use a t-test or an ANOVA/regression model |
No value returned; the file par.file
is simply created in root.dir
.
do.call(make.par.file, extract.pars())
recreates the original parameter file.
See the individual function help pages for each function for more detailed descriptions of the above parameters.
Don Barkauskas (barkda@wald.ucdavis.edu)
Barkauskas, D.A. et al. (2008) “Detecting glycan cancer biomarkers in serum samples using MALDI FT-ICR mass spectrometry data”. Submitted to Bioinformatics
Xi, Y. and Rocke, D.M. (2006) “Baseline Correction for NMR Spectroscopic Metabolomics Data Analysis”, unpublished.