Skip to content

Commit eb0a19e

Browse files
chr1swallacechr1swallace
committed
prepare for 5.2.2 cran release
1 parent c0d3d7f commit eb0a19e

Some content is hidden

Large Commits have some content hidden by default. Use the searchbox below for content that may be hidden.

79 files changed

+4249
-6783
lines changed

Diff for: ‎.travis.yml

-8
This file was deleted.

Diff for: ‎DESCRIPTION

+5-6
Original file line numberDiff line numberDiff line change
@@ -14,11 +14,10 @@ Suggests:
1414
testthat,
1515
mvtnorm,
1616
magrittr,
17-
rmarkdown,
18-
snpStats
17+
rmarkdown
1918
Title: Colocalisation Tests of Two Genetic Traits
20-
Version: 5.2.1
21-
Date: 2022-06-24
19+
Version: 5.2.2
20+
Date: 2023-05-12
2221
Authors@R: c(person("Chris", "Wallace", role=c("aut","cre"),
2322
email = "[email protected]"),
2423
person("Claudia","Giambartolomei", role="aut",
@@ -29,12 +28,12 @@ Maintainer: Chris Wallace <[email protected]>
2928
Description: Performs the colocalisation tests described in
3029
Giambartolomei et al (2013) <doi:10.1371/journal.pgen.1004383>,
3130
Wallace (2020) <doi:10.1371/journal.pgen.1008720>,
32-
Wallace (2021) <doi:10.1101/2021.02.23.432421>.
31+
Wallace (2021) <doi:10.1371/journal.pgen.1009440>.
3332
License: GPL
3433
LazyLoad: yes
3534
VignetteBuilder: knitr
3635
Roxygen: list(markdown = TRUE)
37-
RoxygenNote: 7.2.1
36+
RoxygenNote: 7.2.3
3837
Encoding: UTF-8
3938
URL: https://github.com/chr1swallace/coloc
4039
BugReports: https://github.com/chr1swallace/coloc/issues

Diff for: ‎NEWS.md

+6-3
Original file line numberDiff line numberDiff line change
@@ -1,3 +1,6 @@
1+
# coloc 5.2.2
2+
* better checking of dataset inputs, and better documentation of what that data should look like
3+
14
# coloc 5.2.0
25
* update to new susieR >= 0.12.06 which requires sample size
36
* general tidy up of susie-related code
@@ -70,17 +73,17 @@
7073
# coloc 2.3-2
7174
- tidied code relating mainly to proportional
7275
- colocalisation testing methods, making more methods confirm to S4.
73-
- pcs.prepare now imputates missing genotypes by default
76+
- pcs.prepare now imputes missing genotypes by default
7477
2013-05-22
7578

7679
# coloc 2.3
7780
## coloc BUGFIX
78-
- Introduced a function to estimate trait variance from supplied coefficients and standard errors. This is used within the approach implemented in coloc.abf(), and replaces the earlier version which implicity assumed that var(Y)=1 for quantitative traits, which could lead to incorrect inference when var(Y) was far from 1.
81+
- Introduced a function to estimate trait variance from supplied coefficients and standard errors. This is used within the approach implemented in coloc.abf(), and replaces the earlier version which implicitly assumed that var(Y)=1 for quantitative traits, which could lead to incorrect inference when var(Y) was far from 1.
7982

8083
2013-09-25
8184

8285
# coloc 2.2
83-
- Merged coloc.abf and coloc.abf.imputed(), so that datasets for wheich beta, var(beta) are available can be matched to datasets with only p values and maf.2 This means the arguments to coloc.abf() have been changed! Please
86+
- Merged coloc.abf and coloc.abf.imputed(), so that datasets for which beta, var(beta) are available can be matched to datasets with only p values and maf.2 This means the arguments to coloc.abf() have been changed! Please
8487
check ?coloc.abf for the new function.
8588

8689
2013-19-06

Diff for: ‎R/check.R

+114-106
Original file line numberDiff line numberDiff line change
@@ -1,51 +1,53 @@
11
##' Check coloc dataset inputs for errors
22
##'
3+
##' A coloc dataset is a list, containing a mixture of vectors
4+
##' capturing quantities that vary between snps (these vectors must
5+
##' all have equal length) and scalars capturing quantities that
6+
##' describe the dataset.
7+
##'
38
##' Coloc is flexible, requiring perhaps only p values, or z scores, or effect
49
##' estimates and standard errors, but with this flexibility, also comes
510
##' difficulties describing exactly the combinations of items required.
6-
7-
##' \describe{
811
##'
12+
##' Required vectors are some subset of
13+
##'
14+
##' \describe{
15+
##' \item{beta}{regression coefficient for each SNP from dataset 1}
16+
##' \item{varbeta}{variance of beta}
917
##' \item{pvalues}{P-values for each SNP in dataset 1}
10-
##'
11-
##' \item{N}{Number of samples in dataset 1}
12-
##'
1318
##' \item{MAF}{minor allele frequency of the variants}
19+
##' \item{snp}{a character vector of snp ids, optional. It will be used to merge dataset1 and dataset2 and will be retained in the results.}
20+
##' }
1421
##'
15-
##' \item{beta}{regression coefficient for each SNP from dataset 1}
16-
##'
17-
##' \item{varbeta}{variance of beta}
18-
##'
19-
##' \item{type}{the type of data in dataset 1 - either "quant" or "cc" to denote quantitative or case-control}
20-
##'
21-
##' \item{s}{for a case control dataset, the proportion of samples in dataset 1 that are cases}
22-
##'
23-
##' \item{sdY}{for a quantitative trait, the population standard deviation of the trait. if not given, it can be estimated from the vectors of varbeta and MAF}
24-
##'
25-
##' \item{snp}{a character vector of snp ids, optional. If present, it will be used to merge dataset1 and dataset2. Otherwise, the function assumes dataset1 and dataset2 contain results for the same SNPs in the same order.}
22+
##' Preferably, give \code{beta} and \code{varbeta}. But if these are not available, sufficient statistics can be approximated from \code{pvalues} and \code{MAF}.
23+
##'
24+
##' Required scalars are some subset of
2625
##'
26+
##' \describe{
27+
##' \item{N}{Number of samples in dataset 1}
28+
##' \item{type}{the type of data in dataset 1 - either "quant" or "cc" to denote quantitative or case-control}
29+
##' \item{s}{for a case control dataset, the proportion of samples in dataset 1 that are cases}
30+
##' \item{sdY}{for a quantitative trait, the population standard deviation of the trait. if not given, it can be estimated from the vectors of varbeta and MAF}
2731
##' }
2832
##'
29-
##' Some of these items may be missing, but you must always give {\code{type}}.
30-
##'
31-
##' Then scalars describing the samples used:
33+
##' You must always give {\code{type}}. Then,
3234
##' \describe{
33-
##' \item{always needed}{\code{N}}
3435
##' \item{if \code{type}=="cc"}{\code{s}}
3536
##' \item{if \code{type}=="quant" and \code{sdY} known}{\code{sdY}}
37+
##' \item{if beta, varbeta not known}{\code{N}}
3638
##' }
3739
##' If \code{sdY} is unknown, it will be approximated, and this will require
3840
##' \describe{
3941
##' \item{summary data to estimate \code{sdY}}{\code{beta}, \code{varbeta}, \code{N}, \code{MAF}}
4042
##' }
4143
##'
42-
##' Then, if not already covered above, the summary statistics describing the results
44+
##' Optional vectors are
45+
##'
4346
##' \describe{
44-
##' \item{preferably}{\code{beta}, \code{varbeta}}
45-
##' \item{alternatively}{\code{pvalues}, \code{MAF}}
47+
##' \item{position}{a vector of snp positions, required for \code{plot_dataset}}
4648
##' }
4749
##'
48-
##' \code{check_dataset} call stop() unless a series of expectations on dataset
50+
##' \code{check_dataset} calls stop() unless a series of expectations on dataset
4951
##' input format are met
5052
##'
5153
##' This is a helper function for use by other coloc functions, but
@@ -60,90 +62,96 @@
6062
##' @return NULL if no errors found
6163
##' @export
6264
##' @author Chris Wallace
63-
check_dataset <- function(d,suffix="",req=c("snp"),warn.minp=1e-6) {
64-
if(!is.list(d) )
65-
stop("dataset ",suffix,": is not a list")
66-
nd <- names(d)
67-
68-
## no missing values - make people clean their own data rather than make assumptions here for datasets I don't know
69-
## req <- unique(c("snp",req)) # always need snp to match now
70-
n <- 0
71-
if(length(setdiff(req,nd)))
72-
stop("dataset ",suffix,": missing required element(s) ",paste(setdiff(req,nd),collapse=", "))
73-
74-
for(v in nd) {
75-
if(any(is.na(d[[v]])))
76-
stop("dataset ",suffix,": ",v," contains missing values")
77-
}
78-
79-
## snps should be unique
80-
if("snp" %in% nd && any(duplicated(d$snp)))
81-
stop("dataset ",suffix,": duplicated snps found")
82-
if("snp" %in% nd && is.factor(d$snp))
83-
stop("dataset ",suffix,": snp should be a character vector but is a factor")
84-
85-
## MAF should be > 0, < 1
86-
if("MAF" %in% nd && (!is.numeric(d$MAF) || any(is.na(d$MAF)) ||
87-
any(d$MAF<=0) || any(d$MAF>=1)))
88-
stop("dataset ",suffix,": MAF should be a numeric, strictly >0 & <1")
89-
90-
## lengths of these should match
91-
l <- -1 # impossible length
92-
shouldmatch <- c("pvalues","MAF","beta","varbeta","snp","position")
93-
for(v in shouldmatch)
94-
if(v %in% nd)
95-
if(l<0) { ## update
96-
l <- length(d[[v]])
97-
} else { ## check
98-
if(length(d[[v]])!=l) {
99-
stop("dataset ",suffix,": lengths of inputs don't match: ")
100-
print(intersect(nd, shouldmatch))
65+
check_dataset <- function(d,
66+
suffix="",
67+
req=c("type","snp"),
68+
warn.minp=1e-6) {
69+
if(!is.list(d) )
70+
stop("dataset ",suffix,": is not a list")
71+
recognised_items=c("beta","varbeta","pvalues","MAF","snp","position","N","type","s","sdY","LD")
72+
nd <- intersect(names(d), recognised_items)
73+
74+
## no missing values - make people clean their own data rather
75+
## than make assumptions here for datasets I don't know
76+
n <- 0
77+
if(length(setdiff(req,nd)))
78+
stop("dataset ",suffix,": missing required element(s) ",paste(setdiff(req,nd),collapse=", "))
79+
80+
for(v in nd) {
81+
if(any(is.na(d[[v]])))
82+
stop("dataset ",suffix,": ",v," contains missing values")
83+
}
84+
85+
if (!(d$type %in% c("quant", "cc")))
86+
stop("dataset ", suffix, ": ", "type must be quant or cc")
87+
88+
## snps should be unique
89+
if("snp" %in% nd && any(duplicated(d$snp)))
90+
stop("dataset ",suffix,": duplicated snps found")
91+
if("snp" %in% nd && is.factor(d$snp))
92+
stop("dataset ",suffix,": snp should be a character vector but is a factor")
93+
94+
## MAF should be > 0, < 1
95+
if("MAF" %in% nd && (!is.numeric(d$MAF) || any(is.na(d$MAF)) ||
96+
any(d$MAF<=0) || any(d$MAF>=1)))
97+
stop("dataset ",suffix,": MAF should be a numeric, strictly >0 & <1")
98+
99+
## lengths of vector arguments should match
100+
l <- -1 # impossible length
101+
shouldmatch <- intersect(nd, c("pvalues","MAF","beta","varbeta","snp","position"))
102+
for(v in shouldmatch)
103+
if(l<0) { ## update
104+
l <- length(d[[v]])
105+
} else { ## check
106+
if(length(d[[v]])!=l) {
107+
stop("dataset ",suffix,": lengths of inputs don't match: ")
108+
print(shouldmatch)
109+
}
101110
}
102-
}
103-
104-
## sample size
105-
if (("N" %in% req) && (!('N' %in% nd) || is.null(d$N) || is.na(d$N)))
106-
stop("dataset ",suffix,": sample size N not set")
107-
108-
## type of data
109-
if (! ('type' %in% nd))
110-
stop("dataset ",suffix,": variable type not set")
111-
if(!(d$type %in% c("quant","cc")))
112-
stop("dataset ",suffix,": ","type must be quant or cc")
113-
114-
## no beta/varbeta
115-
if(("s" %in% nd) && (!is.numeric(d$s) || d$s<=0 || d$s>=1))
116-
stop("dataset ",suffix,": ","s must be between 0 and 1")
117-
if(!("beta" %in% nd) || !("varbeta" %in% nd)) { # need to estimate var (Y)
118-
if(!("pvalues" %in% nd) || !( "MAF" %in% nd))
119-
stop("dataset ",suffix,": ","require p values and MAF if beta, varbeta are unavailable")
120-
if(d$type=="cc" && !("s" %in% nd))
121-
stop("dataset ",suffix,": ","require, s, proportion of samples who are cases, if beta, varbeta are unavailable")
122-
p=d$pvalues
123-
} else {
124-
p=pnorm( -abs( d$beta/sqrt(d$varbeta) ) ) * 2
125-
}
126-
127-
## minp
128-
if(min(p) > warn.minp)
129-
warning("minimum p value is: ",format.pval(min(p)),"\nIf this is what you expected, this is not a problem.\nIf this is not as small as you expected, please check the 02_data vignette.")
130-
131-
## sdY
132-
if(d$type=="quant" && !("sdY" %in% nd))
133-
if(!("MAF" %in% nd && "N" %in% nd ))
134-
stop("dataset ",suffix,": ","must give sdY for type quant, or, if sdY unknown, MAF and N so it can be estimated")
135-
136-
if("LD" %in% nd) {
137-
if(nrow(d$LD)!=ncol(d$LD))
138-
stop("LD not square")
139-
if(!identical(colnames(d$LD),rownames(d$LD)))
140-
stop("LD rownames != colnames")
141-
if(length(setdiff(d$snp,colnames(d$LD))))
142-
stop("colnames in LD do not contain all SNPs")
143-
}
144111

145-
## if we reach here, no badness detected
146-
NULL
112+
## type of data
113+
if (! ('type' %in% nd))
114+
stop("dataset ",suffix,": variable type not set")
115+
if(!(d$type %in% c("quant","cc")))
116+
stop("dataset ",suffix,": ","type must be quant or cc")
117+
118+
## no beta/varbeta
119+
if(("s" %in% nd) && (!is.numeric(d$s) || d$s<=0 || d$s>=1))
120+
stop("dataset ",suffix,": ","s must be between 0 and 1")
121+
if(!("beta" %in% nd) || !("varbeta" %in% nd)) { # need to estimate var (Y)
122+
if(!("pvalues" %in% nd) || !( "MAF" %in% nd))
123+
stop("dataset ",suffix,": ","require p values and MAF if beta, varbeta are unavailable")
124+
if(any(d$pvalues<=0))
125+
stop("pvalues should not be negative or exactly 0")
126+
if(d$type=="cc" && !("s" %in% nd))
127+
stop("dataset ",suffix,": ","require, s, proportion of samples who are cases, if beta, varbeta are unavailable")
128+
if (!('N' %in% nd) || is.null(d$N) || any(d$N<=0) )
129+
stop("dataset ",suffix,": sample size N <=0 or not set")
130+
p=d$pvalues
131+
} else {
132+
p=pnorm( -abs( d$beta/sqrt(d$varbeta) ) ) * 2
133+
}
134+
135+
## minp
136+
if(min(p) > warn.minp)
137+
warning("minimum p value is: ",format.pval(min(p)),"\nIf this is what you expected, this is not a problem.\nIf this is not as small as you expected, please check you supplied var(beta) and not sd(beta) for the varbeta argument. If that's not the explanation, please check the 02_data vignette.")
138+
139+
## sdY
140+
if(d$type=="quant" && !("sdY" %in% nd))
141+
if(!("MAF" %in% nd && "N" %in% nd ))
142+
stop("dataset ",suffix,": ","must give sdY for type quant, or, if sdY unknown, MAF and N so it can be estimated")
143+
144+
if("LD" %in% nd) {
145+
if(nrow(d$LD)!=ncol(d$LD))
146+
stop("LD not square")
147+
if(!identical(colnames(d$LD),rownames(d$LD)))
148+
stop("LD rownames != colnames")
149+
if(length(setdiff(d$snp,colnames(d$LD))))
150+
stop("colnames in LD do not contain all SNPs")
151+
}
152+
153+
## if we reach here, no badness detected
154+
NULL
147155
}
148156

149157
#'@rdname check_dataset

0 commit comments

Comments
 (0)