TcGSA.LR | R Documentation |
This function computes the Likelihood Ratios for the gene sets under scrutiny, as well as estimations of genes dynamics inside those gene sets through mixed models.
TcGSA.LR( expr, gmt, design, subject_name = "Patient_ID", time_name = "TimePoint", crossedRandom = FALSE, covariates_fixed = "", time_covariates = "", time_func = "linear", group_name = "", separateSubjects = FALSE, minGSsize = 10, maxGSsize = 500 ) ## S3 method for class 'TcGSA' print(x, ...)
expr |
a matrix or dataframe of gene expression. Its dimension are nxp, with the p samples in column and the n genes in row. |
gmt |
a gmt object containing the gene sets definition. See
|
design |
a matrix or dataframe containing the experimental variables that used in the model,
namely |
subject_name |
the name of the factor variable from |
time_name |
the name of a numeric variable from |
crossedRandom |
logical flag indicating whether the random effects of the subjects and of the time points
should be modeled as one crossed random effect or as two separated random effects.
Default is |
covariates_fixed |
a character vector with the names of numeric or factor variables from the |
time_covariates |
a character vector with the names of numeric or factor variables from the |
time_func |
the form of the time trend. Can be either one of |
group_name |
in the case of several treatment groups, the name of a factor variable
from the |
separateSubjects |
logical flag indicating that the analysis identifies
gene sets that discriminates patients rather than gene sets than have a
significant trend over time. Default is |
minGSsize |
the minimum number of genes in a gene set. If there are
less genes than this number in one of the gene sets under scrutiny, the
Likelihood Ratio of this gene set is not computed (the mixed model are not
fitted). Default is |
maxGSsize |
the maximum number of genes in a gene set. If there are
more genes than this number in one of the gene sets under scrutiny, the
Likelihood Ratio of this gene set is not computed (the mixed model are not
fitted). This is to avoid very long computation times. Default is
|
x |
an object of class ' |
... |
further arguments passed to or from other methods. |
This Time-course Gene Set Analysis aims at identifying gene sets that are not
stable over time, either homogeneously or heterogeneously (see Hejblum
et al, 2012)in terms of their probes. And when the argument
separateSubjects
is TRUE
, instead of identifying gene sets that
have a significant trend over time, TcGSA identifies gene sets that
have significantly different trends over time depending on Subjects
.
TcGSA.LR
returns a tcgsa
object, which is a list with
the 5 following elements:
fit a data frame that contains the 3 following variables:
LR
: the likelihood ratio between the model under the
null hypothesis and the model under the alternative hypothesis.
CVG_H0
: convergence status of the model under the null hypothesis.
CVG_H1
: convergence status of the model under the alternative
hypothesis.
time_func
: a character string passing along the value of the
time_func
argument used in the call.
GeneSets_gmt
: a gmt
object passing along the value of the
gmt
argument used in the call.
group.var
: a factor passing along the group_name
variable
from the design
matrix.
separateSubjects
: a logical flag passing along the value of the
separateSubjects
argument used in the call.
Estimations
: a list of 3 dimensions arrays. Each element of the
list (i.e. each array) corresponds to the estimations of gene expression
dynamics for each of the gene sets under scrutiny (obtained from mixed
models). The first dimension of those arrays is the genes included in the
concerned gene set, the second dimension is the Patient_ID
, and the
third dimension is the TimePoint
. The values inside those arrays are
estimated gene expressions.
time_DF
: the degree of freedom of the natural splines functions
Boris P. Hejblum
Hejblum BP, Skinner J, Thiebaut R, (2015) Time-Course Gene Set Analysis for Longitudinal Gene Expression Data. PLOS Comput. Biol. 11(6):e1004310. doi: 10.1371/journal.pcbi.1004310
summary.TcGSA
, plot.TcGSA
,
and TcGSA.LR.parallel
for an implementation using
parallel computing
if(interactive()){ data(data_simu_TcGSA) tcgsa_sim_1grp <- TcGSA.LR(expr=expr_1grp, gmt=gmt_sim, design=design, subject_name="Patient_ID", time_name="TimePoint", time_func="linear", crossedRandom=FALSE) tcgsa_sim_1grp summary(tcgsa_sim_1grp) plot(x=tcgsa_sim_1grp, expr=expr_1grp, Subject_ID=design$Patient_ID, TimePoint=design$TimePoint, baseline=1, B=100, time_unit="H" ) tcgsa_sim_2grp <- TcGSA.LR(expr=expr_2grp, gmt=gmt_sim, design=design, subject_name="Patient_ID", time_name="TimePoint", time_func="linear", crossedRandom=FALSE, group_name="group.var") tcgsa_sim_2grp }
Add the following code to your website.
For more information on customizing the embed code, read Embedding Snippets.