svdCMAP: Calculate and remove principal components from Connectivity...

Description Usage Arguments Value Examples

View source: R/svdCMAP.R

Description

reconstructCMAP: calculate and remove principal components from Connectivity map data. If CMAP data matrix have already been reconstructed using the same PCs and pert_types this function will read previously saved SVD output and reconstructed CMAP data matrix

svdCMAP: calculate and remove principal components from Connectivity map data. If CMAP data matrix have already been reconstructed using the same PCs and pert_types this function will read previously saved SVD output and reconstructed CMAP data matrix

Usage

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
reconstructCMAP(CMap_files, remove_PCs = "none", return_PCs = T,
  return_CMAP = T, save_CMAP = F, pert_types = c("trt_sh.cgs",
  "trt_oe", "trt_cp", "trt_lig", "trt_oe.mut", "trt_xpr", "trt_sh.css",
  "ctl_vehicle.cns", "ctl_vector.cns", "ctl_untrt.cns"))

svdCMAP(CMap_files, return_PCs = T, is_touchstone = "all",
  pert_types = c("trt_sh.cgs", "trt_oe", "trt_cp", "trt_lig",
  "trt_oe.mut", "trt_xpr", "trt_sh.css", "ctl_vehicle.cns",
  "ctl_vector.cns", "ctl_untrt.cns"), pert_itimes = c("all"),
  cell_ids = c("all"), gene_names = c("all"), keep_one_oe = "all",
  landmark_only = F, ...)

Arguments

CMap_files

a list of directories and urls produced by loadCMap

remove_PCs

integer or character, indexes - which principal components to remove. "none" to keep all PCs

return_PCs

logical, return the output of svd decomposition

return_CMAP

logical, return Connectivity map data reconstructed using only specified PCs

save_CMAP

logical, save Connectivity map data reconstructed using only specified PCs. Returned CMap_files contain the path to reconstructed Connectivity map data matrix and corresponding supplementary files.

pert_types

a character vector of perturbation types, query for available perturbation types using perturbTable(CMap_files, ~ pert_type). Details: https://docs.google.com/document/d/1q2gciWRhVCAAnlvF2iRLuJ7whrGP6QjpsCMq1yWz7dU/edit#

is_touchstone

logical, select only the perturbation data that is a part of the touchstone dataset (genes profiled across all 9 core cell lines)? Details: https://docs.google.com/document/d/1q2gciWRhVCAAnlvF2iRLuJ7whrGP6QjpsCMq1yWz7dU/

pert_itimes

a character vector of perturbation times, query for available perturbation times using perturbTable(CMap_files, ~ pert_itime)

cell_ids

a character vector of cell line names, query for available cell line names using perturbTable(CMap_files, ~ cell_id)

gene_names

a character vector of HUGO gene names. Use "all" to select all perturbations.

keep_one_oe

keep only one overexpression experiment per gene and condition. Applicable only when pert_types = "trt_oe". Perturbations where sig_id matches pert_id are retained ("one"). To invert the selection use "other". To select all use "all"

landmark_only

read only landmark genes (rows, measured). Details: https://docs.google.com/document/d/1q2gciWRhVCAAnlvF2iRLuJ7whrGP6QjpsCMq1yWz7dU/edit

...

passed to svd: nu and nv

CMap_files

a list of directories and urls produced by loadCMap

return_PCs

logical, return the output of svd decomposition. Alternatively just calculate and save.

pert_types

a character vector of perturbation types, query for available perturbation types using perturbTable(CMap_files, ~ pert_type). Details: https://docs.google.com/document/d/1q2gciWRhVCAAnlvF2iRLuJ7whrGP6QjpsCMq1yWz7dU/edit#

Value

reconstructCMAP:

svdCMAP:

Examples

1
2
3
4
library(regNETcmap)
CMap_files = loadCMap(directory = "../regulatory_networks_by_cmap/data/cmap/")
library(regNETcmap)
CMap_files = loadCMap(directory = "../regulatory_networks_by_cmap/data/cmap/")

vitkl/regNETcmap documentation built on Feb. 18, 2020, 3:43 a.m.