cTWAS analysis using summary statistics

ctwas_sumstats(
  z_snp,
  weights,
  region_info,
  LD_map,
  snp_map,
  z_gene = NULL,
  thin = 0.1,
  niter_prefit = 3,
  niter = 30,
  L = 5,
  init_group_prior = NULL,
  init_group_prior_var = NULL,
  group_prior_var_structure = c("shared_type", "shared_context", "shared_nonSNP",
    "shared_all", "independent"),
  filter_L = TRUE,
  filter_nonSNP_PIP = FALSE,
  min_nonSNP_PIP = 0.5,
  min_p_single_effect = 0.8,
  maxSNP = Inf,
  min_var = 2,
  min_gene = 1,
  min_group_size = 100,
  use_null_weight = TRUE,
  coverage = 0.95,
  min_abs_corr = 0.1,
  LD_format = c("rds", "rdata", "mtx", "csv", "txt", "custom"),
  LD_loader_fun = NULL,
  snpinfo_loader_fun = NULL,
  force_compute_cor = FALSE,
  save_cor = FALSE,
  cor_dir = NULL,
  outputdir = NULL,
  outname = "ctwas",
  ncore = 1,
  ncore_LD = max(ncore - 1, 1),
  seed = 99,
  logfile = NULL,
  verbose = FALSE,
  ...
)

Arguments

z_snp: A data frame with four columns: "id", "A1", "A2", "z". giving the z scores for SNPs. "A1" is effect allele. "A2" is the other allele.
weights: a list of pre-processed prediction weights.
region_info: a data frame of region definitions.
LD_map: a data frame with filenames of LD matrices and SNP information for the regions.
snp_map: a list of data frames with SNP-to-region map for the reference.
z_gene: A data frame with columns: "id", "z", giving the z-scores for genes.
thin: The proportion of SNPs to be used for estimating parameters and screening regions.
niter_prefit: the number of iterations of the E-M algorithm to perform during the initial parameter estimation step.
niter: the number of iterations of the E-M algorithm to perform during the complete parameter estimation step.
L: the number of effects for susie during the fine mapping steps.
init_group_prior: a vector of initial values of prior inclusion probabilities for SNPs and genes.
init_group_prior_var: a vector of initial values of prior variances for SNPs and gene effects.
group_prior_var_structure: a string indicating the structure to put on the prior variance parameters. "shared_type" allows all groups in one molecular QTL type to share the same variance parameter. "shared_context" allows all groups in one context (tissue, cell type, condition) to share the same variance parameter. "shared_nonSNP" allows all non-SNP groups to share the same variance parameter. "shared_all" allows all groups to share the same variance parameter. "independent" allows all groups to have their own separate variance parameters.
filter_L: If TRUE, screening regions with L > 0.
filter_nonSNP_PIP: If TRUE, screening regions with total non-SNP PIP >= min_nonSNP_PIP.
min_nonSNP_PIP: Regions with non-SNP PIP >= min_nonSNP_PIP will be selected to run finemapping using all SNPs.
min_p_single_effect: Regions with probability greater than min_p_single_effect of having 1 or fewer effects will be used for parameter estimation.
maxSNP: Inf or integer. Maximum number of SNPs in a region. Default is Inf, no limit. This can be useful if there are many SNPs in a region and you don't have enough memory to run the program.
min_var: minimum number of variables (SNPs and genes) in a region when estimating paramters and screening regions.
min_gene: minimum number of genes in a region when estimating paramters and screening regions.
min_group_size: Minimum number of genes in a group. Groups with number of genes < min_group_size will be removed for the analysis.
use_null_weight: If TRUE, allow for a probability of no effect in susie.
coverage: A number between 0 and 1 specifying the “coverage” of the estimated confidence sets.
min_abs_corr: Minimum absolute correlation allowed in a credible set.
LD_format: file format for LD matrix. If "custom", use a user defined LD_loader_fun() function to load LD matrix.
LD_loader_fun: a user defined function to load LD matrix when LD_format = "custom".
snpinfo_loader_fun: a user defined function to load SNP information file, if SNP information files are not in standard cTWAS reference format.
force_compute_cor: If TRUE, force computing correlation (R) matrices.
save_cor: If TRUE, save correlation (R) matrices to cor_dir.
cor_dir: The directory to store correlation (R) matrices.
outputdir: The directory to store output. If specified, save outputs to the directory.
outname: The output name.
ncore: The number of cores used to parallelize computing over regions.
ncore_LD: The number of cores used to parallelize computing correlation matrices, in screening regions and fine-mapping steps with LD.
seed: seed for random sampling when thinning the SNPs in region data.
logfile: The log filename. If NULL, print log info on screen.
verbose: If TRUE, print detailed messages.
...: Additional arguments of susie_rss.

Value

a list, including z_gene, estimated parameters, region_data, cross-boundary genes, screening region results, and fine-mapping results.