make_threshold: Determine best cutoff point of CMap similarity score.

Description Usage Arguments Author(s) References Examples

View source: R/make_threshold.R

Description

This function calculate the best cutoff point of CMap score by using LINCS information file and user-determined true perturbation IDs. Before using this function, users must run init function only one time. Users set the output directory of init function as input directory of this function.

Usage

1
make_threshold(pert_id_vec, cscore, cell, input.dir, output.dir)

Arguments

pert_id_vec

character vector of pert_id, described in LINCS inst.info file. LINCS data containing this pert_id vector are considered as true hit in ROC analysis.

cscore

a object from cscore_calc function.

cell

character vector of cell_line, described in LINCS inst.info file naming a directory containing the result of init function.

input.dir

a character string naming a directory containing the result of init function.

output.dir

a character string naming a directory for output files from this function.

Author(s)

Masaru Koido (masaru.koido@jfcr.or.jp)

References

M. Koido, et al., 'Detection of hub genes for developing gene expression networks under anticancer drug treatment', Oncotarget, 9(49), 29097-29111, 2018.

http://www.lincsproject.org/

Examples

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15
16
## Not run:
#-- only the first time of InDePTH
# init( input.dir = "input", out.dir = "init" )
#
# Run
res1 <- cscore_LINCS(up.sig = c( "217761_at", "217398_x_at", "218744_s_at" ),
                     dn.sig = c( "221856_s_at", "218193_s_at", "218457_s_at" ),
                     input.dir = "init",
                     output.dir = "output",
                     write.name = NULL)
make_threshold( pert_id_vec = c( "TRCN0000010389", "TRCN0000010390", "TRCN0000010391"),
                cscore = res1,
                cell = "HT29",
                input.dir = "init",
                output.dir = "output" )
## End(Not run)

koido/InDePTH documentation built on May 28, 2019, 9:50 a.m.