MRCIEU · remlapmot · Sep 5, 2023 · Aug 9, 2023 · Aug 9, 2023 · Aug 9, 2023
diff --git a/DESCRIPTION b/DESCRIPTION
@@ -1,11 +1,13 @@
 Package: gwasvcf
 Title: Tools for Dealing with GWAS Summary Data in VCF Format
-Version: 0.1.1
+Version: 0.1.2
 Authors@R: c(
     person("Gibran", "Hemani", , "g.hemani@bristol.ac.uk", role = c("aut", "cre"),
            comment = c(ORCID = "0000-0003-0920-1055")),
     person("Tom", "Palmer", , "tom.palmer@bristol.ac.uk", role = "ctb",
-           comment = c(ORCID = "0000-0003-4655-4511"))
+           comment = c(ORCID = "0000-0003-4655-4511")),
+    person("Rita", "Rasteiro", , "rita.rasteiro@bristol.ac.uk", role = "ctb",
+           comment = c(ORCID = "0000-0002-4217-3060"))
   )
 Description: Tools for dealing with GWAS summary data in VCF format.
     Includes reading, querying, writing, as well as helper functions such
@@ -23,6 +25,7 @@ Imports:
     genetics.binaRies,
     GenomeInfoDb,
     GenomicRanges,
+    gwasglue2,
     IRanges,
     magrittr,
     RCurl,
@@ -41,7 +44,9 @@ Suggests:
 VignetteBuilder: 
     knitr
 Remotes:
-    github::mrcieu/genetics.binaRies
+    github::mrcieu/genetics.binaRies,
+    github::mrcieu/gwasglue2
 Encoding: UTF-8
-RoxygenNote: 7.2.2
+Roxygen: list(markdown = TRUE)
+RoxygenNote: 7.2.3
 SystemRequirements: GNU unzip
diff --git a/NAMESPACE b/NAMESPACE
@@ -9,6 +9,7 @@ export(create_rsidx_index_from_vcf)
 export(create_rsidx_sub_index)
 export(create_vcf)
 export(get_ld_proxies)
+export(gwasvcf_to_summaryset)
 export(merge_vcf)
 export(parse_chrompos)
 export(proxy_match)

diff --git a/NEWS.md b/NEWS.md
@@ -0,0 +1,4 @@
+# gwasvcf 0.1.2
+
+* New `gwasvcf_to_summaryset()` function to create a [gwasglue2](https://mrcieu.github.io/gwasglue2) SummarySet object from a vcf file
+* Fixed error in `get_ld_proxies()` related with argument `validate`, deprecated in `as_tibble()` (tibble 2.0.0)
diff --git a/R/gwasglue.R b/R/gwasglue.R
@@ -0,0 +1,21 @@
+#  This file contains the functions to create a gwasglue2 SummarySet object
+
+
+
+#' Create a SummarySet
+#' 
+#' Returns a gwasglue2 SummarySet object
+#' @param vcf Path or URL to GWAS-VCF file or VCF object e.g. output from [VariantAnnotation::readVcf()], [create_vcf()] or [query_gwas()]
+#' @export
+gwasvcf_to_summaryset <- function(vcf){
+	# get metadata from vcf and create metadata object
+	md <- gwasglue2::create_metadata(id = vcf@metadata$header@samples, build = unique(VariantAnnotation::meta(header(vcf))$contig$assembly))
+
+	# get summary data and create SummarySet
+
+    s <- vcf %>% 
+		vcf_to_tibble() %>% 
+		gwasglue2::create_summaryset_from_gwasvcf(metadata = md)
+
+    return(s)
+}
diff --git a/R/manipulate.r b/R/manipulate.r
@@ -191,7 +191,7 @@ vcf_to_tibble <- function(vcf, id=NULL)
 		return(dplyr::tibble())
 	}
 	S4Vectors::values(a)[["rsid"]] <- names(a)
-	return(dplyr::as_tibble(a))
+	return(dplyr::as_tibble(a, .name_repair = "minimal"))
 }
 
 

diff --git a/R/proxy.r b/R/proxy.r
@@ -56,7 +56,7 @@ get_ld_proxies <- function(rsid, bfile, searchspace=NULL, tag_kb=5000, tag_nsnp=
 		return(ld)
 	}
 	ld <- data.table::fread(paste0("gunzip -c ", outname), header=TRUE) %>%
-		dplyr::as_tibble() %>%
+		dplyr::as_tibble(.name_repair="minimal") %>%
 		dplyr::filter(.data[["R"]]^2 > tag_r2) %>%
 		dplyr::filter(.data[["SNP_A"]] != .data[["SNP_B"]]) %>%
 		dplyr::mutate(PHASE=gsub("/", "", .data[["PHASE"]])) %>%
@@ -70,9 +70,9 @@ get_ld_proxies <- function(rsid, bfile, searchspace=NULL, tag_kb=5000, tag_nsnp=
 		message("No proxies found")
 		return(ld)
 	}
-	temp <- do.call(rbind, strsplit(ld[["PHASE"]], "")) %>% dplyr::as_tibble(.data, .name_repair="minimal")
+	temp <- do.call(rbind, strsplit(ld[["PHASE"]], "")) %>% dplyr::as_tibble(.name_repair="minimal")
 	names(temp) <- c("A1", "B1", "A2", "B2")
-	ld <- cbind(ld, temp) %>% dplyr::as_tibble(.data, .name_repair="minimal")
+	ld <- cbind(ld, temp) %>% dplyr::as_tibble(.name_repair="minimal")
 	# ld <- dplyr::arrange(ld, desc(abs(R))) %>%
 	# 	dplyr::filter(!duplicated(SNP_A))
 	ld <- dplyr::arrange(ld, dplyr::desc(abs(.data[["R"]])))
@@ -99,16 +99,16 @@ sqlite_ld_proxies <- function(rsids, dbfile, tag_r2)
 	numid <- gsub("rs", "", rsids) %>% paste(collapse=",")
 	query <- paste0("SELECT DISTINCT * FROM tags WHERE SNP_A IN (", numid, ")")
 	ld <- RSQLite::dbGetQuery(conn, query) %>% 
-		dplyr::as_tibble() %>%
+		dplyr::as_tibble(.name_repair="minimal") %>%
 		dplyr::filter(.data[["R"]]^2 > tag_r2) %>%
 		dplyr::filter(.data[["SNP_A"]] != .data[["SNP_B"]]) %>%
 		dplyr::mutate(PHASE=gsub("/", "", .data[["PHASE"]])) %>%
 		dplyr::filter(nchar(.data[["PHASE"]]) == 4) %>%
 		dplyr::mutate(SNP_A = paste0("rs", .data[["SNP_A"]]), SNP_B = paste0("rs", .data[["SNP_B"]]))
 
-	temp <- do.call(rbind, strsplit(ld[["PHASE"]], "")) %>% dplyr::as_tibble(.data, .name_repair="minimal")
+	temp <- do.call(rbind, strsplit(ld[["PHASE"]], "")) %>% dplyr::as_tibble(.name_repair="minimal")
 	names(temp) <- c("A1", "B1", "A2", "B2")
-	ld <- cbind(ld, temp) %>% dplyr::as_tibble(.data, .name_repair="minimal")
+	ld <- cbind(ld, temp) %>% dplyr::as_tibble(.name_repair="minimal")
 	ld <- dplyr::arrange(ld, dplyr::desc(abs(.data[["R"]])))
 	message("Found ", nrow(ld), " proxies")
 	RSQLite::dbDisconnect(conn)

diff --git a/R/pval_index.r b/R/pval_index.r
@@ -1,6 +1,6 @@
 #' Create pval index from GWAS-VCF file
 #'
-#' Create a separate file called <id>.pvali which is used to speed up p-value queries
+#' Create a separate file called `<id>.pvali` which is used to speed up p-value queries.
 #'
 #' @param vcffile VCF filename
 #' @param maximum_pval Maximum p-value to include. Default = 0.05

diff --git a/R/query.r b/R/query.r
@@ -2,7 +2,7 @@
 #'
 #' Read in GWAS summary data with filters on datasets (if multiple datasets per file) and/or chromosome/position, rsids or pvalues. Chooses most optimal choice for the detected operating system. Typically chrompos searches are the fastest. On Windows, rsid or pvalue filters from a file will be slow. 
 #'
-#' @param vcf Path or URL to GWAS-VCF file or VCF object e.g. output from VariantAnnotation::readVcf or gwasvcftools::query_vcf
+#' @param vcf Path or URL to GWAS-VCF file or VCF object e.g. output from [VariantAnnotation::readVcf()] or [create_vcf()]
 #' @param chrompos Either vector of chromosome and position ranges e.g. "1:1000" or "1:1000-2000", or data frame with columns `chrom`, `start`, `end`.
 #' @param rsid Vector of rsids
 #' @param pval  P-value threshold (NOT -log10)

diff --git a/man/create_pval_index_from_vcf.Rd b/man/create_pval_index_from_vcf.Rd
diff --git a/man/gwasvcf_to_summaryset.Rd b/man/gwasvcf_to_summaryset.Rd
diff --git a/man/parse_chrompos.Rd b/man/parse_chrompos.Rd
diff --git a/man/query_chrompos_bcftools.Rd b/man/query_chrompos_bcftools.Rd
diff --git a/man/query_chrompos_file.Rd b/man/query_chrompos_file.Rd
diff --git a/man/query_chrompos_vcf.Rd b/man/query_chrompos_vcf.Rd
diff --git a/man/query_gwas.Rd b/man/query_gwas.Rd
diff --git a/man/vcflist_overlaps.Rd b/man/vcflist_overlaps.Rd
diff --git a/vignettes/guide.Rmd b/vignettes/guide.Rmd
@@ -342,3 +342,19 @@ writeVcf(out, file="temp.vcf")
 ```
 
 You may want to first harmonise the data so that all the non-effect alleles are aligned to the human genome reference. See the [gwasglue](https://github.com/MRCIEU/gwasglue) package on some functions to do this. 
+
+## Creating a gwasglue2 SummarySet object from a vcf file
+
+Although still under development, if compared with its predecessor, the [gwasglue2](https://mrcieu.github.io/gwasglue2/) package has several new features, including the use of S4 R objects.
+
+It is possible to create a `SummarySet` object from a GWAS-VCF file or VCF object e.g. output from `VariantAnnotation::readVcf()`, `create_vcf()` or `query_gwas()` using the `gwasvcf_to_summaryset()` function.
+
+
+For example:
+
+```{r, eval=FALSE}
+summaryset <- readVcf(vcffile) %>% 
+              gwasvcf_to_summaryset()
+```
+
+Once the `SummarySet` objects are created, it is possible to use `gwasglue2` to harmonise data, harmonise against a LD matrix, remap genomic coordinates to a different genome assembly, convert to other formats and more.