Bioconductor fixes

.Rbuildignore

@@ -0,0 +1,11 @@
+PTMsToPathways.Rproj
+.github/
+data-raw/
+pkgdown/
+docs/
+_pkgdown.yml
+scripts/
+^vignettes/.*\.html$
+^vignettes/.*\.txt$
+^vignettes/.*\.Rout$
+^vignettes/plots/

.gitignore

@@ -1,14 +1,20 @@
-.Rbuildignore
 .Rproj.user
 .Rhistory
 .RData
 .temp
 *.DS_Store
 /PTMsToPathways_0.0.0.9000/
-PTMsToPathways_0.0.0.9000.tar.gz
+PTMsToPathways*.tar.gz
 
 README.html
 
+vignettes/*.html
+vignettes/*.txt
+vignettes/*.Rout
+vignettes/plots/
+
 docs/
 /Meta/
 inst/doc
+
+*.Rcheck

DESCRIPTION

@@ -1,15 +1,23 @@
 Package: PTMsToPathways
-Title: Filtered CoCluster Correlation Network Guide
+Title: Post-Translational Modification Network and Pathway Analysis
 Version: 0.99.0
 Authors@R: c(
-    person("Mark", "Grimes", email = "first.last@example.com", role = c("aut", "cre")),
-    person("Mikhail", "Ukrainetz", email = "first.last@example.com", role = c("aut")),
-    person("Grant", "Smith", email = "first.last@example.com", role = c("aut")),
-    person("Madison", "Moffett", email = "first.last@example.com", role = c("aut")),
-    person("Nagashree", "Avabhrath", email = "first.last@example.com", role = c("aut")),
-    person("Lucia", "Williams", email = "first.last@example.com", role = c("aut"))
+    person("Mark", "Grimes", role = c("aut"), comment=c(ORCID = "0000-0003-2673-5892")),
+    person("Mikhail", "Ukrainetz", role = c("aut")),
+    person("Grant", "Smith",  role = c("aut")),
+    person("Madison", "Moffett", role = c("aut")),
+    person("Nagashree", "Avabhrath", role = c("aut")),
+    person("Lucia", "Williams", email = "lucia.williams@umt.edu", role = c("aut", "cre"), comment=c(ORCID = "0000-0003-3785-0247"))
            )
-Description: A collection of functions in order to process mass spectrometry data of post-translational modifications into possible pathways.
+Description: A collection of functions to integrate PTM proteomic data with
+ protein-protein interactions and cell signaling pathways.
+ These functions can be used to cluster PTMs to reveal patterns
+ in their expression under different experimental conditions. These clustering
+ patterns can then used to filter protein-protein interactions from external PPI
+ databases and pathway-pathway interactions. The resulting networks are
+ navigable data structures that reveal data-supported interactions among both
+ individual proteins and groups of proteins representing well-defined cell
+ signaling pathways.
 License: GPL (>= 3)
 Encoding: UTF-8
 Roxygen: list(markdown = TRUE)
@@ -29,18 +37,16 @@ Imports:
     ggplot2,
     rlang
 Suggests: 
-    BiocManager,
     knitr,
     rmarkdown,
     testthat (>= 3.0.0)
 VignetteBuilder:
     knitr
 Depends: 
     R (>= 4.5.0)
-URL: https://um-applied-algorithms-lab.github.io/PTMsToPathways/
+URL: https://um-applied-algorithms-lab.github.io/PTMsToPathways/, https://github.com/UM-Applied-Algorithms-Lab/PTMsToPathways
 BugReports: https://github.com/UM-Applied-Algorithms-Lab/PTMsToPathways/issues/new
 LazyData: true
 LazyDataCompression: xz
 Config/testthat/edition: 3
-Additional_repositories: https://bioconductor.org/packages/3.21/bioc
 biocViews: Software, Proteomics, PostTranslationalModifications, Pathways, Network

NAMESPACE

@@ -40,7 +40,4 @@ export(setNodeColorToRowz)
 export(setNodeMapping)
 export(setNodeSizeColorIndependently)
 export(strip.cy.goo)
-importFrom(gplots,heatmap.2)
-importFrom(grDevices,colorRampPalette)
-importFrom(grDevices,hcl)
 importFrom(rlang,.data)

R/BRCAHelpers.R

@@ -59,9 +59,6 @@
 #' )
 #' }
 #'
-#' @importFrom gplots heatmap.2
-#' @importFrom grDevices hcl colorRampPalette
-#'
 graph.ptm.by.cluster <- function(
     ptmtable,
     common.clusters,
@@ -174,8 +171,8 @@ graph.ptm.by.cluster <- function(
   ## image() requires: nrow(z) == length(x)-1, ncol(z) == length(y)-1
   write_heat_key <- function(filename, zlim = 3) {
     n     <- 256L                                     # number of colour cells
-    cols  <- colorRampPalette(c("#3333FF","#E6E6E6","#FFFF00"),
-                               space = "rgb", interpolate = "linear")(n)
+    cols  <- grDevices::colorRampPalette(c("#3333FF", "#E6E6E6", "#FFFF00"),
+                                         space = "rgb", interpolate = "linear")(n)
     # breakpoints: length n+1 so the z matrix is (1 x n)
     brks  <- seq(-zlim, zlim, length.out = n + 1L)
     z     <- matrix(seq_len(n), nrow = 1L, ncol = n)  # 1 x n
@@ -240,8 +237,8 @@ graph.ptm.by.cluster <- function(
 
   # ---- colour setup --------------------------------------------------
 
-  heatcols        <- colorRampPalette(c("#3333FF","#E6E6E6","#FFFF00"),
-                                      space="rgb", interpolate="linear")(255)
+  heatcols        <- grDevices::colorRampPalette(c("#3333FF", "#E6E6E6", "#FFFF00"),
+                                                 space = "rgb", interpolate = "linear")(255)
   breaks          <- seq(-zlim, zlim, length.out = 256)
 
   cluster.palette <- make_cluster_colors(length(ordered.blocks))

R/PTMsToPathways-package.R

@@ -0,0 +1,19 @@
+#' PTMsToPathways: Filtered CoCluster Correlation Network Guide
+#'
+#' PTMsToPathways provides tools for turning mass spectrometry data on
+#' post-translational modifications into clustered networks, co-cluster
+#' correlation networks, and downstream pathway and Cytoscape-ready views.
+#' The package includes example PTM tables, BRCA example networks, and helper
+#' functions for building CCCNs, cluster-filtered PPIs, and pathway crosstalk
+#' networks.
+#'
+#' Typical entry points are [MakeClusterList()], [MakeCorrelationNetwork()],
+#' [BuildClusterFilteredNetwork()], [BuildPathwayCrosstalkNetwork()], and the
+#' Cytoscape helpers in [GraphCfn()] and related functions.
+#'
+#' The package comes with many example datasets that can be seen in the vignettes. 
+#'
+#' @name PTMsToPathways
+#' @keywords package
+#' @seealso [utils::browseVignettes()]
+"_PACKAGE"

_pkgdown.yml

@@ -16,7 +16,7 @@ navbar:
         - text: "Raw Data Processing"
           href: articles/RawDataProcessing.html
         - text: "Creating Networks"
-          href: articles/CreatingNetworks.html
+          href: articles/GettingStarted.html
         - text: "Cytoscape Graphing"
           href: articles/CytoscapeGraphing.html
         - text: "BRCA Networks"
@@ -27,6 +27,6 @@ articles:
   navbar: ~
   contents:
   - RawDataProcessing
-  - CreatingNetworks
+  - GettingStarted
   - CytoscapeGraphing
   - BRCANetworks

data-raw/ex-creating-networks-data.R

@@ -1,4 +1,4 @@
-# create the full example data as in the vignette CreatingNetworks.
+# create the full example data as in the getting started vignette/
 # The randomness inherent in the clustering procedure, and the MakeClusterList parameters keep length
 # and toolong, make it so that the data generated here may be slightly different than the existing example
 # data, which would break tests. So we don't overwrite the existing data, but provide the code for reference.

inst/scripts/README.md

@@ -0,0 +1,50 @@
+# PTMsToPathways data provenance
+
+This directory documents how package data and bundled extdata files are
+prepared.
+
+## Source-of-truth scripts
+
+- `data-raw/function-key.R` reads `inst/extdata/FunctionKey.txt` and writes
+  `data/function_key.rda`.
+- `data-raw/ex-pathways-list.R` reads the BioPlanet pathway table and writes
+  `data/ex_pathways_list.rda`.
+- `data-raw/ex-creating-networks-data.R` derives the example clustering and
+  CCCN objects used in the Getting Started vignette.
+- `data-raw/full-PCN-edgelist.R` documents the larger example PCN workflow.
+- `data-raw/processed-brca.R` derives the BRCA example objects from the BRCA
+  phosphoproteomics table and local STRING/GeneMANIA/BioPlanet inputs.
+
+## Bundled extdata sources
+
+- `inst/extdata/AlldataPTMs.txt` is the larger PTM example used in the
+  Getting Started vignette. The source URL is documented in the vignette and
+  points to the package repository copy of the file.
+- `inst/extdata/pathway.csv` is a BioPlanet pathway table used by
+  `ReadBioplanetFile()`; the public source is
+  https://tripod.nih.gov/bioplanet/download/pathway.csv.
+- `inst/extdata/bioplanet_pathway_June2025.csv` is a package-local BioPlanet
+  snapshot used for vignette and data generation workflows.
+- `inst/extdata/PhosphoDiscoData_mmc9.txt` is the BRCA phosphoproteomics input
+  table used in the BRCA vignette and `data-raw/processed-brca.R`.
+- `inst/extdata/PhosDiscoModules_mmc11.txt` is the associated module annotation
+  table used in the BRCA vignette.
+- `inst/extdata/FunctionKey.txt` is the annotation key used to generate
+  `function_key`.
+- `inst/extdata/small_string_hs_hugo.tsv` is the preprocessed local STRING
+  example used to keep vignette and build-time examples offline. It is a subset
+  of the dataset available on [Zenodo](https://zenodo.org/record/1234567).
+- `inst/extdata/Kinase_Substrate_Dataset.txt` is the kinase-substrate example
+  input used by `GetKinsub.edges()`.
+- `inst/extdata/ex_genemania_interactions.txt` and `inst/extdata/ex_gm_results.txt`
+  are example GeneMANIA outputs used in the vignettes and tests.
+
+## Regeneration notes
+
+- The BRCA example objects are derived from the BRCA phosphoproteomics input,
+  then filtered through `MakeClusterList()`, `MakeCorrelationNetwork()`,
+  `GetSTRINGdb.edges()`, `GetGeneMANIA.edges()`, and
+  `BuildPathwayCrosstalkNetwork()`.
+- Several generated data objects rely on randomized clustering. The example
+  scripts preserve the existing package data rather than overwriting it when the
+  stochastic outputs would differ.

vignettes/BRCANetworks.Rmd

@@ -129,7 +129,7 @@ head(ptmtable[, 1:5])
 ### Create Clusters and Co-Cluster Correlation Networks (CCCNs)
 
 Next, we create clusters and networks from those clusters as in the
-[Creating Networks vignette](CreatingNetworks.html). This takes about 10
+[Creating Networks vignette](GettingStarted.html). This takes about 10
 minutes on a laptop, so we provide both the code and the pre-computed
 results for this step. To re-run the analysis, run, the following:
 
@@ -156,7 +156,7 @@ adj.consensus.matrix <- clusterlist.data[[2]]
 ptm.correlation.matrix <- clusterlist.data[[3]]
 ```
 
-These are required for the next step.
+These are required to create the PTM and Gene co-cluster correlation networks.
 
 ```{r eval = TRUE}
 CCCN.data <- MakeCorrelationNetwork(adj.consensus.matrix,
@@ -238,9 +238,10 @@ eval_brca[1:10, ]
 ### Build Cluster Filtered Networks (CFNs) and Pathway Crosstalk Networks (PCNs)
 
 For PPI edges, the code below demonstrates how to get the STRING-db and
-GeneMANIA edges from the static human PPI data downloaded as local files. [todo: instructions
-to download once these are available.] Alternatively, the PPI data
-can be obtained from STRINGdb and GeneMANIA websites as demonstrated in the [Creating Networks vignette](CreatingNetworks.html).
+GeneMANIA edges from the static human PPI data downloaded as local files. These
+files are available on [Zenodo](https://zenodo.org/records/20631767).
+Alternatively, the PPI data
+can be obtained from STRINGdb and GeneMANIA websites as demonstrated in the [getting started with P2P vignette](GettingStarted.html).
 
 ```{r eval = FALSE}
 string_db_filepath <- "your/filepath/here.tsv"
@@ -271,7 +272,7 @@ head(stringdb.edges)
 The GeneMANIA human PPI edge file contains the following types of interactions: 
 "Genetic Interactions", "Pathway", "Physical Interactions", and "Predicted."
 
-We choose all but "Genetic Interactions" to include using gm.interaction.types in the following function. 
+We choose all but "Genetic Interactions" to include using the `gm.interaction.types` parameter in the following function. 
 
 ```{r eval = FALSE}
 genemania_db_filepath <- "your/filepath/here.tsv"
@@ -291,13 +292,17 @@ genemania.edges <- BRCA_genemania.edges
 head(genemania.edges)
 ```
 
-Next, we retrieve kinase-substrate edges, then obain the cluster filtered network, retaining PPIs only for proteins whose PTMs co-cluster, as demonstrated in the [Creating Networks vignette](CreatingNetworks.html).
+Next, we retrieve kinase-substrate edges, then obain the cluster filtered
+network, retaining PPIs only for proteins whose PTMs co-cluster, as demonstrated
+in the [getting started vignette](GettingStarted.html).
 
 ```{r eval = TRUE}
 file_path <- system.file("extdata", "Kinase_Substrate_Dataset.txt", package = "PTMsToPathways")
 kinsub.edges <- GetKinsub.edges(file_path, gene.cccn.nodes)
+```
 
-# Now we can build the CFN.
+Now we can build the CFN.
+```{r eval = TRUE}
 network.list <- BuildClusterFilteredNetwork(gene.cccn.edges,
                                             stringdb.edges,
                                             genemania.edges,
@@ -350,7 +355,8 @@ dim(PD_module.df) # should be 1017 rows
 head(PD_module.df)
 ```
 
-We note that there are differences from the PTM table imported above. We will work with those sites that match between ptmtable and PD_module.df
+We note that there are differences from the PTM table imported above. We will
+work with those sites that match between `ptmtable` and `PD_module.df`.
 
 ```{r eval = TRUE}
 length(intersect(newphos$variable_sites_names, PD_module.df$variableSites)) # 530
@@ -413,18 +419,19 @@ length(mod63.clust.ptms)
 ```
 
 P2P provides functions to prepare visualizations of these PTMs in
-Cytoscape. This code graphs the CFN/CCCN from all these PTMs:
+[Cytoscape](https://cytoscape.org/). First, we create a dataframe of
+node information that can be used with `RCy3` functions to create a network in Cytoscape.
 
 ```{r eval = TRUE}
- funckey <- function_key
+funckey <- function_key
 cfn.cccn <- ptms_to_cfn(mod63.clust.ptms, cfn = cfn.merged, ptm.cccn.edges = ptm.cccn.edges, ptmtable = ptmtable, pepsep = ";")
 cfn_cccn.nodes <- make.cytoscape.node.file(cfn.cccn, funckey, ptmtable,
                                            include.gene.data = TRUE,
                                            include.coclustered.PTMs = TRUE,
                                            ptm.cccn.edges = ptm.cccn.edges)
 ```
 
-To graph in Cytoscape:
+To graph in Cytoscape, use the P2P function `GraphCfn`:
 
 ```{r eval = FALSE}
 g1 <- GraphCfn(cfn.edges = cfn.cccn, cfn.nodes = cfn_cccn.nodes,
@@ -459,13 +466,12 @@ cfn_cccn.nodes.cdksubs <- cfn_cccn.nodes[cfn_cccn.nodes$Gene.Name %in% cdk1.subs
 ```
 
 Two ways to select for CDK1 substrates are presented. Method 1: Using
-RCy3.
+`RCy3`.
 
 ```{r eval = FALSE}
-library(RCy3)
-selectNodes(cfn_cccn.nodes.cdksubs$id, by = "id", preserve=FALSE)
-selectEdgesConnectingSelectedNodes()
-createSubnetwork(nodes = getSelectedNodes(), edges = getSelectedEdges(), nodes.by.col = "id", edges.by.col = "name")
+RCy3::selectNodes(cfn_cccn.nodes.cdksubs$id, by = "id", preserve=FALSE)
+Cy3::selectEdgesConnectingSelectedNodes()
+RCy3::createSubnetwork(nodes = RCy3::getSelectedNodes(), edges = RCy3::getSelectedEdges(), nodes.by.col = "id", edges.by.col = "name")
 ```
 
 Method 2: Using P2P functions.
@@ -510,7 +516,6 @@ mutant); X21BR010 (PIK3CA missense mutant); and X05BR045 (TP53 nonsense
 and MLLT4 frameshift mutants).
 
 ```{r eval = FALSE}
-library(RCy3)
 setNodeColorToRatios(plotcol="X03BR011")
 setNodeColorToRowz(plotcol="X03BR011")  # This exaggerates the node size and shape somewhat. 
 setNodeColorToRatios(plotcol="X21BR010")
@@ -522,3 +527,9 @@ setNodeColorToRowz(plotcol="X05BR045")
 Note that different samples have dramatically different differences in
 PTMs that are up or down, which is reflected also in total in gene
 nodes.
+
+### Session Info
+
+```{r eval = TRUE}
+sessionInfo()
+```