v1.37.1 - updates following removal of library spectra from msPurityData package (too large for Bioconductor) (#108)

* Use zenodo reference to library spectra

* Update CI/CD

* troubleshoot vignette build issue

* Troubleshooting cache on github workflows

* Troubleshooting cache for github workflows

* more troubleshooting GH workflow cache

Author: Tomnl

.github/workflows/check-bioc.yml

@@ -52,9 +52,9 @@ jobs:
       fail-fast: false
       matrix:
         config:
-          - { os: ubuntu-latest, r: 'devel', bioc: 'devel', cont: "bioconductor/bioconductor_docker:devel", rspm: "https://packagemanager.rstudio.com/cran/__linux__/focal/latest" }
+          - { os: ubuntu-latest, r: 'devel', bioc: 'devel', cont: "bioconductor/bioconductor_docker:devel", rspm: "https://packagemanager.rstudio.com/cran/__linux__/jammy/latest" }
           #- { os: macOS-latest, r: 'next', bioc: '3.19'}
-          - { os: windows-latest, r: 'next', bioc: '3.19'}
+          # - { os: windows-latest, r: 'next', bioc: '3.22'}
           ## Check https://github.com/r-lib/actions/tree/master/examples
           ## for examples using the http-user-agent
     env:

.travis.yml

@@ -2,8 +2,8 @@ Package: msPurity
 Type: Package
 Title: Automated Evaluation of Precursor Ion Purity for Mass Spectrometry Based
     Fragmentation in Metabolomics
-Version: 1.31.1
-Date: 2024-05-09
+Version: 1.37.1
+Date: 2025-02-19
 Authors@R: c(
     person(given = "Thomas N.", family = "Lawson",
         email = "thomas.nigel.lawson@gmail.com", role = c("aut", "cre"),
@@ -48,7 +48,8 @@ Imports:
     fastcluster,
     ggplot2,
     DBI,
-    RSQLite
+    RSQLite,
+    BiocFileCache
 Suggests:
     MSnbase,
     testthat,
@@ -61,7 +62,7 @@ Suggests:
     RPostgres,
     RMySQL
 VignetteBuilder: knitr
-RoxygenNote: 7.3.1
+RoxygenNote: 7.3.3
 Roxygen: list(markdown = TRUE)
 biocViews: MassSpectrometry, Metabolomics, Software
 Collate: 

DESCRIPTION

@@ -1,3 +1,9 @@
+CHANGES IN VERSION 1.37.1
++ Removed library_spectra.db direct reference from msPurityData package (too large for Bioconductor)
++ spectralMatching() now downloads default library database from Zenodo (https://zenodo.org/records/18700802) and uses BiocFileCache with MD5 verification
++ spectral_matching() (deprecated function) now requires explicit library_db_pth parameter; default library no longer available
++ Added BiocFileCache to Imports for automatic caching of remote databases 
+
 CHANGES IN VERSION 1.30.1 / 1.31.1
 + Update example for purityX causing build error
 

NEWS

@@ -21,14 +21,13 @@
 #'
 #' \href{https://bioconductor.org/packages/release/bioc/html/msPurity.html}{msPurity Bioconductor}
 #'
-#' @docType package
-#' @name msPurity
+#' @keywords internal
 #' @importFrom magrittr %>%
 #' @importFrom grDevices dev.off png
 #' @importFrom graphics abline legend lines plot points text
 #' @importFrom stats approxfun dnorm median na.omit sd
 #' @importFrom utils fix head packageVersion read.csv write.csv
-NULL
+"_PACKAGE"
 
 ## quiets concerns of R CMD check re: the .'s that appear in pipelines
 if(getRversion() >= "2.15.1")  utils::globalVariables(c("."))

R/msPurity.R

@@ -84,7 +84,7 @@ spectral_matching <- function(query_db_pth, ra_thres_l=0, ra_thres_q=2, cores=1,
                                      pa=NA, xset=NA, grp_peaklist=NA, out_dir='.', ra_w=0.5, mz_w=2,
                                      spectra_type_q="scans", ra_thres_t=NA, target_db_pth=NA, rt_range=c(NA, NA), rttol=NA,
                                      match_alg='dpc'){
-  message("Running msPurity spectral matching function for LC-MS(/MS) data")
+  message("Running msPurity spectral matching function for LC-MS(/MS) data [this function is deprecated, please use msPurity::spectralMatching for future use]")
 
   if (!is.na(ra_thres_t)){
     message("ra_thres_t argument has been deprecated and will be remove in future versions of msPurity,
@@ -108,7 +108,7 @@ spectral_matching <- function(query_db_pth, ra_thres_l=0, ra_thres_q=2, cores=1,
   }
 
   if (is.na(library_db_pth)){
-    library_db_pth <- system.file("extdata", "library_spectra", "library_spectra.db", package="msPurityData")
+    stop("Default library database is no longer bundled. Download from https://zenodo.org/records/18700802/files/library_spectra.db?download=1 and pass via library_db_pth.")
   }
 
 

R/spectral-matching.R

@@ -288,7 +288,30 @@ spectralMatching <- function(
 
 
   if (is.na(l_dbPth)){
-    l_dbPth <- system.file("extdata", "library_spectra", "library_spectra.db", package="msPurityData")
+
+    bfc <- BiocFileCache::BiocFileCache(ask = FALSE)
+    cache_name <- "msPurity_library_spectra_db"
+    cache_url <- "https://zenodo.org/records/18700802/files/library_spectra.db?download=1"
+    expected_md5 <- "f6a4033d376278cccfc30d1911fb2fa5"
+    cache_hit <- BiocFileCache::bfcquery(bfc, cache_name, "rname", exact = TRUE)
+
+    # Prefer existing cache entry to avoid re-downloading large data.
+    if (nrow(cache_hit) == 0){
+      # Add to cache if not present (download occurs on first use).
+      BiocFileCache::bfcadd(bfc, cache_name, cache_url)
+      cache_hit <- BiocFileCache::bfcquery(bfc, cache_name, "rname", exact = TRUE)
+    }
+
+    if (nrow(cache_hit) == 0){
+      stop("Unable to create or locate BiocFileCache entry for the default library database.")
+    }
+
+    rid <- cache_hit$rid[1]
+    l_dbPth <- BiocFileCache::bfcrpath(bfc, rids = rid)
+    actual_md5 <- tolower(unname(tools::md5sum(l_dbPth)))
+    if (!identical(actual_md5, expected_md5)){
+      stop("Downloaded library database failed MD5 verification.")
+    }
   }
 
   ########################################################