A function to understand the novelties within CSeQTL. The user can experiment with trimming TReC, assess parameter estimation, perform hypothesis testing, actively constrain cell type-specific parameters, when analyzing a single gene/SNP pair.
CSeQTL_smart(
TREC,
hap2,
ASREC,
PHASE,
SNP,
RHO,
XX,
upPHI,
upKAPPA,
upETA,
upPSI,
upALPHA,
iFullModel = FALSE,
trim = FALSE,
thres_TRIM = 10,
hypotest = TRUE,
swap = TRUE,
numAS = 5,
numASn = 5,
numAS_het = 5,
cistrans_thres = 0.01,
gr_eps = 0.01,
conv_eps = 0.001,
ncores = 1,
show = FALSE
)A R list of statistics and metrics after optimizing the model for a single SNP. The list contains parameter MLEs, gradient vectors, covariance matrices, gene expression estimates, fold change estimates, convergence indicators, likelihood ratio test statistics, and associated p-values.
An integer vector of total read counts.
An integer vector of second haplotype counts
An integer vector of total haplotype counts
An integer vector of 0s and 1s denoting if a subject has available haplotype counts.
An integer vector of phased genotypes coded 0 (AA), 1 (AB), 2 (BA), 3 (BB), and 5 (NA).
A numeric matrix of cell type proportions. Rows correspond to subjects and columns correspond to cell types.
A numeric design matrix of baseline covariates including the intercept in the first column and centered continuous covariates.
A value of 0 or 1 indicating if a Poisson or Negative binomial distribution is fitted, respectively.
An integer vector of zeroes and ones where
length(upKAPPA) == ncol(RHO). A requirement is upKAPPA[1] = 1.
Cell types with indices equal to one have the baseline fold change
between the q-th and reference cell type are estimated. Otherwise
that cell type's parameter is constrained to zero.
An integer vector of zeroes and ones where
length(upETA) == ncol(RHO) indicating which
cell types eQTL parameters are estimated or constrained to their null.
A value of 0 or 1 indicating if a Binomial or Beta-binomial distribution is fitted, respectively.
An integer vector of zeroes and ones where
length(upALPHA) == ncol(RHO) indicating which
cell types' cis-trans parameters are estimated or
constrained to their null.
A boolean that if set to TRUE will
determine the submodel of the full model, based by upKAPPA,
upETA, and upALPHA parameters, that can be estimated with stability.
Boolean value. If TRUE, the TReC model will be fitted
without SNP genotype to calculate each subject's Cooks' distance.
A positive numeric value to perform subject outcome trimming. Subjects with standardized Cooks' Distances greater than the threshold are trimmed.
A boolean to perform eQTL significance testing and cis-trans eQTL testing.
A boolean to determine if the reference cell type should be swapped with the cell type with highest TReC across alleles.
A positive integer to determine if a subject has enough total haplotype counts.
A positive integer to determine how many subjects
have at least numAS to use the haplotype counts.
A positive integer to determine how many subjects
with at least numAS are heterozygous (AB or BA). If
sum(PHASE == 1 & ASREC >= numAS & (SNP == 1 | SNP == 2)) >= numAS_het,
those subjects haplotype counts will be used for TReCASE and cis/trans
testing and estimation.
A numeric value to determine the cis/trans test p-value cutoff.
A numeric value to determine if convergence is achieved based on the L2 norm of the gradient.
A numeric value to determine if convergence is achieved based on the L2 norm of the product of the inverse hessian and gradient.
A positive integer specifying the number of threads available to decrease computational runtime.
A boolean value to display verbose output and plot intermediate simulated results.