-
Notifications
You must be signed in to change notification settings - Fork 8
/
spia.r
57 lines (43 loc) · 1.78 KB
/
spia.r
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
record2pathway = function(rec, exp, lookup) {
spia = import('../../util/spia')
b = import('base/operators')
print(rec)
rownames(exp) = lookup$entrezgene[match(rownames(exp), lookup$hgnc_symbol)]
exp = limma::avereps(exp[!is.na(rownames(exp)),])
result = spia$spia(exp[,rec$perturbed], exp[,rec$control], pathids=spia$speed2kegg)
# # if per_sample=TRUE
# colnames(result) = spia$kegg2speed[colnames(result)]
# result
setNames(result, spia$kegg2speed[names(result)])
}
library(dplyr)
import('base/operators')
io = import('io')
ar = import('array')
hpc = import('hpc')
EXPR = commandArgs(TRUE)[1] %or% '../../data/expr.RData'
OUTFILE = commandArgs(TRUE)[2] %or% "spia.RData"
# get index, expr data for test set
speed = io$load(EXPR)
# HGNC -> entrez gene lookup
lookup = biomaRt::useMart(biomart="ensembl", dataset="hsapiens_gene_ensembl") %>%
biomaRt::getBM(attributes=c("hgnc_symbol", "entrezgene"), mart=.)
#scores = mapply(record2pathway, rec=index, exp=expr,
# MoreArgs=list(lookup=lookup), SIMPLIFY=FALSE)
result = hpc$Q(record2pathway,
rec = speed$records, exp = speed$expr,
const = list(lookup = lookup),
memory=4096, n_jobs=10, fail_on_error=FALSE) %>%
setNames(names(speed$records))
errors = sapply(result, function(r) class(r) == "try-error")
if (any(errors)) {
print(result[errors])
result[errors] = NA
}
scores = ar$stack(result, along=1) %>%
ar$map(along=1, scale) %>% # normalize different magnitude of pathways
ar$map(along=2, scale) # normalize total activation per experiment
filter_index = function(x) x[! names(x) %in% c('control', 'perturbed', 'exclusion')]
index = lapply(speed$records[rownames(scores)], filter_index) %>%
do.call(bind_rows, .)
save(scores, index, file=OUTFILE)