Browse code

scone_1.1.3

mbcole authored on 19/10/2017 20:52:03
Showing1 changed files
... ...
@@ -24,6 +24,9 @@
24 24
 #'
25 25
 #' @importFrom RColorBrewer brewer.pal
26 26
 #' @importFrom rARPACK svds
27
+#' @importFrom graphics boxplot
28
+#' @importFrom utils write.csv
29
+#' @importFrom stats setNames var
27 30
 #' @export
28 31
 #'
29 32
 #' @return An object that represents the SCONE report app.
... ...
@@ -66,15 +69,15 @@ sconeReport = function(x, methods,
66 69
     stop("reshape2 package needed for sconeReport()")
67 70
   }
68 71
 
69
-  if (!require("plotly", quietly = TRUE)) {
72
+  if (!requireNamespace("plotly", quietly = TRUE)) {
70 73
     stop("plotly package needed for sconeReport()")
71 74
   }
72 75
 
73
-  if (!require("visNetwork", quietly = TRUE)) {
76
+  if (!requireNamespace("visNetwork", quietly = TRUE)) {
74 77
     stop("visNetwork package needed for sconeReport()")
75 78
   }
76 79
 
77
-  if (!require("ggplot2", quietly = TRUE)) {
80
+  if (!requireNamespace("ggplot2", quietly = TRUE)) {
78 81
     stop("ggplot2 package needed for sconeReport()")
79 82
   }
80 83
 
... ...
@@ -223,18 +226,18 @@ sconeReport = function(x, methods,
223 226
                             ),
224 227
                             shiny::tabPanel("PCA",
225 228
                                      shiny::br(),
226
-                                     shiny::p(paste0("This panel shows principal ",
229
+                                shiny::p(paste0("This panel shows principal ",
227 230
                                               "component analysis ",
228 231
                                               "(PCA) results",
229 232
                                               " on different subsets ",
230 233
                                               "of genes.")),
231 234
                                      shiny::br(),
232
-                                     shiny::h6("Variance Explained (All Genes)"),
235
+                                  shiny::h6("Variance Explained (All Genes)"),
233 236
                                      shiny::p(paste0("Use this plot",
234 237
                                               " to decide on the ",
235 238
                                               "dimensionality of the reduced",
236 239
                                               " spaced used for evaluation.")),
237
-                                     shiny::plotOutput("plot_scree",width = "650px",
240
+                                shiny::plotOutput("plot_scree",width = "650px",
238 241
                                                 height = "400px"),
239 242
                                      shiny::br(),
240 243
                                      shiny::h6("2D (All Genes)"),
... ...
@@ -264,7 +267,7 @@ sconeReport = function(x, methods,
264 267
                             ),
265 268
                             shiny::tabPanel("QC",
266 269
                                      shiny::br(),
267
-                                     shiny::p(paste0("This panel shows the absolute",
270
+                              shiny::p(paste0("This panel shows the absolute",
268 271
                                               " correlations between",
269 272
                                               " Principal",
270 273
                                               " Components (PCs) of",
... ...
@@ -326,7 +329,7 @@ sconeReport = function(x, methods,
326 329
                             ),
327 330
                             shiny::tabPanel("Control Genes",
328 331
                                      shiny::br(),
329
-                                     shiny::p(paste0("Heatmap of control genes, ",
332
+                                  shiny::p(paste0("Heatmap of control genes, ",
330 333
                                               "colored by all",
331 334
                                               " three categories.")),
332 335
                                      shiny::br(),
Browse code

Fix bug in shiny app

Davide Risso authored on 16/10/2017 15:17:51
Showing1 changed files
... ...
@@ -625,7 +625,7 @@ sconeReport = function(x, methods,
625 625
              fill = cc[sort(unique(factor(strat_col())))])
626 626
     })
627 627
 
628
-    output$plot3d_base <- shiny::renderPlotly({
628
+    output$plot3d_base <- plotly::renderPlotly({
629 629
       PC1 <- PC2 <- PC3 <- NULL
630 630
       df <- setNames(data.frame(pc_obj_base()$x[,1:3]),
631 631
                      c("PC1", "PC2", "PC3"))
... ...
@@ -665,7 +665,7 @@ sconeReport = function(x, methods,
665 665
       abs(cor(qc,pc_obj_base()$x))
666 666
     })
667 667
 
668
-    output$qccorPlot <- shiny::renderPlotly({
668
+    output$qccorPlot <- plotly::renderPlotly({
669 669
       metric <- value <- PC <- NULL
670 670
 
671 671
       df = reshape2::melt(cor_qc()[,1:input$dim])
... ...
@@ -696,7 +696,7 @@ sconeReport = function(x, methods,
696 696
              fill = cc[sort(unique(factor(strat_col())))])
697 697
     })
698 698
 
699
-    output$plot3d_qc <- shiny::renderPlotly({
699
+    output$plot3d_qc <- plotly::renderPlotly({
700 700
       if(ncol(pc_obj_qc()$x) >= 3){
701 701
         PC1 <- PC2 <- PC3 <- NULL
702 702
         df <- setNames(data.frame(pc_obj_qc()$x[,1:3]),
Browse code

Shiny app packages to Suggests:

Davide Risso authored on 30/06/2017 21:47:09
Showing1 changed files
... ...
@@ -1,8 +1,8 @@
1 1
 #' SCONE Report Browser: Browse Evaluation of Normalization Performance
2
-#' 
3
-#' This function opens a shiny application session for visualizing performance 
2
+#'
3
+#' This function opens a shiny application session for visualizing performance
4 4
 #' of a variety of normalization schemes.
5
-#' 
5
+#'
6 6
 #' @param x a \code{SconeExperiment} object
7 7
 #' @param methods character specifying the normalizations to report.
8 8
 #' @param qc matrix. QC metrics to be used for QC evaluation report. Required.
... ...
@@ -10,43 +10,31 @@
10 10
 #'   Default NULL.
11 11
 #' @param batch factor. A known batch variable (variation to be removed).
12 12
 #'   Default NULL.
13
-#' @param negcon character. Genes to be used as negative controls for 
14
-#'   evaluation. These genes should be expected not to change according to the 
13
+#' @param negcon character. Genes to be used as negative controls for
14
+#'   evaluation. These genes should be expected not to change according to the
15 15
 #'   biological phenomenon of interest. Default empty character.
16
-#' @param poscon character. Genes to be used as positive controls for 
17
-#'   evaluation. These genes should be expected to change according to the 
16
+#' @param poscon character. Genes to be used as positive controls for
17
+#'   evaluation. These genes should be expected to change according to the
18 18
 #'   biological phenomenon of interest. Default empty character.
19
-#' @param eval_proj function. Projection function for evaluation  (see 
20
-#'   \code{\link{score_matrix}} for details). If NULL, PCA is used for 
19
+#' @param eval_proj function. Projection function for evaluation  (see
20
+#'   \code{\link{score_matrix}} for details). If NULL, PCA is used for
21 21
 #'   projection.
22 22
 #' @param eval_proj_args list. List of args passed to projection function as
23 23
 #'   eval_proj_args.
24
-#'   
25
-#' @importFrom shiny fluidPage downloadLink br titlePanel tableOutput helpText
26
-#'   downloadHandler column h5 h6 updateSelectInput plotOutput tabsetPanel
27
-#'   selectInput renderTable p sidebarLayout sidebarPanel shinyApp mainPanel
28
-#'   observeEvent renderPlot reactive fluidRow sliderInput tabPanel
29
-#' @import visNetwork
30
-#' @importFrom DT dataTableProxy dataTableOutput renderDataTable selectRows
31
-#'   formatSignif
32
-#' @import plotly
24
+#'
33 25
 #' @importFrom RColorBrewer brewer.pal
34
-#' @importFrom reshape2 melt
35
-#' @importFrom NMF aheatmap
36
-#' @importFrom ggplot2 labs theme geom_point ylim ggplot geom_violin
37
-#'   element_blank aes geom_bar coord_cartesian scale_fill_manual guides
38 26
 #' @importFrom rARPACK svds
39 27
 #' @export
40
-#' 
28
+#'
41 29
 #' @return An object that represents the SCONE report app.
42
-#'   
30
+#'
43 31
 #' @examples
44 32
 #' set.seed(101)
45 33
 #' mat <- matrix(rpois(1000, lambda = 5), ncol=10)
46 34
 #' colnames(mat) <- paste("X", 1:ncol(mat), sep="")
47 35
 #' obj <- SconeExperiment(mat)
48 36
 #' res <- scone(obj, scaling=list(none=identity, uq=UQ_FN, deseq=DESEQ_FN),
49
-#'            evaluate=TRUE, k_ruv=0, k_qc=0, eval_kclust=2, 
37
+#'            evaluate=TRUE, k_ruv=0, k_qc=0, eval_kclust=2,
50 38
 #'            bpparam = BiocParallel::SerialParam())
51 39
 #' qc = as.matrix(cbind(colSums(mat),colSums(mat > 0)))
52 40
 #' rownames(qc) = colnames(mat)
... ...
@@ -54,47 +42,74 @@
54 42
 #' \dontrun{
55 43
 #' sconeReport(res,rownames(get_params(res)), qc = qc)
56 44
 #' }
57
-#' 
45
+#'
58 46
 sconeReport = function(x, methods,
59
-                       qc, 
47
+                       qc,
60 48
                        bio = NULL, batch = NULL,
61 49
                        poscon = character(), negcon = character(),
62 50
                        eval_proj = NULL,
63 51
                        eval_proj_args = NULL){
64
-  
65
-  
52
+
53
+  if (!requireNamespace("shiny", quietly = TRUE)) {
54
+    stop("shiny package needed for sconeReport()")
55
+  }
56
+
57
+  if (!requireNamespace("DT", quietly = TRUE)) {
58
+    stop("DT package needed for sconeReport()")
59
+  }
60
+
61
+  if (!requireNamespace("NMF", quietly = TRUE)) {
62
+    stop("NMF package needed for sconeReport()")
63
+  }
64
+
65
+  if (!requireNamespace("reshape2", quietly = TRUE)) {
66
+    stop("reshape2 package needed for sconeReport()")
67
+  }
68
+
69
+  if (!require("plotly", quietly = TRUE)) {
70
+    stop("plotly package needed for sconeReport()")
71
+  }
72
+
73
+  if (!require("visNetwork", quietly = TRUE)) {
74
+    stop("visNetwork package needed for sconeReport()")
75
+  }
76
+
77
+  if (!require("ggplot2", quietly = TRUE)) {
78
+    stop("ggplot2 package needed for sconeReport()")
79
+  }
80
+
66 81
   scone_res = list()
67
-  
82
+
68 83
   # List of normalized matrices
69
-  scone_res$normalized_data = lapply(as.list(methods), 
84
+  scone_res$normalized_data = lapply(as.list(methods),
70 85
                                      FUN = function(z){
71 86
                                        get_normalized(x,method = z,log=TRUE)
72 87
                                      })
73 88
   names(scone_res$normalized_data) = methods
74
-  
89
+
75 90
   # Parameter matrix
76 91
   scone_res$params = get_params(x)[methods,]
77
-  
92
+
78 93
   # Merged score matrix
79 94
   scone_res$scores = cbind(get_scores(x),get_score_ranks(x))[methods,]
80 95
   colnames(scone_res$scores)[ncol(scone_res$scores)] = "mean_score_rank"
81
-  
82
-  
96
+
97
+
83 98
   ## ----- If NULL classifications, Replace with NA ------
84
-  
99
+
85 100
   if(is.null(bio)){
86 101
     bio = factor(rep("NA",ncol(scone_res$normalized_data[[1]])))
87 102
   }
88
-  
103
+
89 104
   if(is.null(batch)){
90 105
     batch = factor(rep("NA",ncol(scone_res$normalized_data[[1]])))
91 106
   }
92
-  
107
+
93 108
   ## ----- Network Data for Visualization -----
94
-  
109
+
95 110
   # Matrix nodes in scone_res
96
-  leaves = rownames(scone_res$params) 
97
-  
111
+  leaves = rownames(scone_res$params)
112
+
98 113
   # Parents
99 114
   lay1 = unique(gsub(",[^,]*$","",leaves))
100 115
   lay2 = unique(gsub(",[^,]*$","",lay1))
... ...
@@ -102,37 +117,37 @@ sconeReport = function(x, methods,
102 117
   lay4 = unique(gsub(",[^,]*$","",lay3))
103 118
   all_nodes = c(leaves,lay1,lay2,lay3,lay4)
104 119
   rm(list = c("lay1","lay2","lay3","lay4"))
105
-  
120
+
106 121
   # Collapse no-op names for all node names
107 122
   all_nodes = unique(gsub("^,*|,*$","",
108 123
                           gsub("\\,+",",",
109 124
                                gsub("none|no_uv|no_bio|no_batch","",
110 125
                                     all_nodes))))
111
-  
126
+
112 127
   # Order by decreasing number of operations (root at the end)
113 128
   all_nodes = all_nodes[order(paste0(gsub(".+",",",all_nodes),
114 129
                                      gsub("[^,]*","",all_nodes)),
115 130
                               decreasing = TRUE)]
116
-  
131
+
117 132
   # Collapse no-op names for all leaves
118 133
   leaves = gsub("^,*|,*$","",
119 134
                 gsub("\\,+",",",
120 135
                      gsub("none|no_uv|no_bio|no_batch","",
121 136
                           leaves)))
122
-  
137
+
123 138
   # Re-title nodes in scone_res and root (instead of collapsed names)
124 139
   titles = all_nodes
125 140
   names(titles) =  paste0("_",all_nodes)
126 141
   titles[paste0("_",leaves)] = rownames(scone_res$params)
127 142
   titles[titles == ""] = "none"
128
-  
143
+
129 144
   # Create Nodes
130 145
   nodes <- data.frame(id = 1:length(all_nodes),
131 146
                       title = titles,
132 147
                       group = c("NoData",
133
-                                "Loaded")[1 + 
148
+                                "Loaded")[1 +
134 149
                                             (all_nodes  %in% leaves)])
135
-  
150
+
136 151
   # Create Edges
137 152
   edges = data.frame()
138 153
   for(i in 1:length(all_nodes)){
... ...
@@ -146,7 +161,7 @@ sconeReport = function(x, methods,
146 161
     }
147 162
   }
148 163
   colnames(edges)= c("from","to")
149
-  
164
+
150 165
   # Re-map indices according to performance
151 166
   new_title_list = rownames(scone_res$params) # Ordered by perfomance
152 167
   new_title_list = c(new_title_list,
... ...
@@ -166,90 +181,90 @@ sconeReport = function(x, methods,
166 181
     edges[old_edges == old_ids[i]] = i
167 182
   }
168 183
   nodes = nodes[order(nodes$id),]
169
-  
184
+
170 185
   ## ----- Color Variable -----
171 186
   cc <- c(brewer.pal(9, "Set1"), brewer.pal(8, "Set2"),brewer.pal(12,"Set3"),
172 187
           brewer.pal(8, "Accent"))
173
-  
188
+
174 189
   ## ----- UI Definition -----
175
-  ui <- fluidPage(
176
-    titlePanel("SCONE Report Browser"),
177
-    
178
-    sidebarLayout(
179
-      sidebarPanel(
180
-        selectInput("norm_code", label = "Normalization",
190
+  ui <- shiny::fluidPage(
191
+    shiny::titlePanel("SCONE Report Browser"),
192
+
193
+    shiny::sidebarLayout(
194
+      shiny::sidebarPanel(
195
+        shiny::selectInput("norm_code", label = "Normalization",
181 196
                     choices = structure(as.list(rownames(scone_res$params)),
182 197
                                         names = rownames(scone_res$params)),
183 198
                     selected = rownames(scone_res$params)[1]),
184
-        selectInput("color_code", label = "Stratify Plots by",
199
+        shiny::selectInput("color_code", label = "Stratify Plots by",
185 200
                     choices = list("Batch"="batch",
186
-                                   "Biological Condition"="bio", 
201
+                                   "Biological Condition"="bio",
187 202
                                    "PAM Cluster"="clust"),
188 203
                     selected = "batch"),
189
-        sliderInput("dim", label = "Reduced Dimension (PCA)",
204
+        shiny::sliderInput("dim", label = "Reduced Dimension (PCA)",
190 205
                     min=2, max=min(length(batch)-1,10),
191 206
                     value = 3),
192
-        sliderInput("k", label = "Number of Clusters (PAM)",
207
+        shiny::sliderInput("k", label = "Number of Clusters (PAM)",
193 208
                     min=2, max=10,
194 209
                     value = 2),
195
-        helpText(paste0("Please note that with many cells (>500),",
210
+        shiny::helpText(paste0("Please note that with many cells (>500),",
196 211
                         " some plots may take several seconds ",
197 212
                         "(sometimes minutes) to appear.")),
198
-        downloadLink('downloadData', 'Download')
213
+        shiny::downloadLink('downloadData', 'Download')
199 214
       ),
200
-      
201
-      mainPanel(tabsetPanel(type = "tabs",
202
-                            tabPanel("Overview",
203
-                                     br(),
215
+
216
+      shiny::mainPanel(shiny::tabsetPanel(type = "tabs",
217
+                            shiny::tabPanel("Overview",
218
+                                     shiny::br(),
204 219
                                      visNetworkOutput("norm_net",
205 220
                                                       width = "650px"),
206
-                                     br(),
221
+                                     shiny::br(),
207 222
                                      DT::dataTableOutput('tbl_score')
208 223
                             ),
209
-                            tabPanel("PCA",
210
-                                     br(),
211
-                                     p(paste0("This panel shows principal ",
224
+                            shiny::tabPanel("PCA",
225
+                                     shiny::br(),
226
+                                     shiny::p(paste0("This panel shows principal ",
212 227
                                               "component analysis ",
213 228
                                               "(PCA) results",
214 229
                                               " on different subsets ",
215 230
                                               "of genes.")),
216
-                                     br(),
217
-                                     h6("Variance Explained (All Genes)"),
218
-                                     p(paste0("Use this plot",
231
+                                     shiny::br(),
232
+                                     shiny::h6("Variance Explained (All Genes)"),
233
+                                     shiny::p(paste0("Use this plot",
219 234
                                               " to decide on the ",
220 235
                                               "dimensionality of the reduced",
221 236
                                               " spaced used for evaluation.")),
222
-                                     plotOutput("plot_scree",width = "650px",
237
+                                     shiny::plotOutput("plot_scree",width = "650px",
223 238
                                                 height = "400px"),
224
-                                     br(),
225
-                                     h6("2D (All Genes)"),
226
-                                     plotOutput("plot_base",
239
+                                     shiny::br(),
240
+                                     shiny::h6("2D (All Genes)"),
241
+                                     shiny::plotOutput("plot_base",
227 242
                                                 width = "650px",
228 243
                                                 height = "450px"),
229
-                                     br(),
230
-                                     h6("3D Interactive (All Genes)"),
244
+                                     shiny::br(),
245
+                                     shiny::h6("3D Interactive (All Genes)"),
231 246
                                      plotlyOutput("plot3d_base",
232 247
                                                   width = "650px",
233 248
                                                   height = "650px"),
234
-                                     br(),
235
-                                     selectInput("gene_set", 
249
+                                     shiny::br(),
250
+                                     shiny::selectInput("gene_set",
236 251
                                                  label = "Gene selection",
237
-                                                 choices = 
252
+                                                 choices =
238 253
                                                    list("Negative Controls"=
239 254
                                                           "neg",
240 255
                                                         "Positive Controls"=
241 256
                                                           "pos"),
242 257
                                                  selected = "neg"),
243
-                                     h6("2D (Select Genes)"),
244
-                                     p(paste0("Visualize cells by PCs",
258
+                                     shiny::h6("2D (Select Genes)"),
259
+                                     shiny::p(paste0("Visualize cells by PCs",
245 260
                                               " of control gene sets.")),
246
-                                     plotOutput("plot_select",
261
+                                     shiny::plotOutput("plot_select",
247 262
                                                 width = "650px",
248 263
                                                 height = "450px")
249 264
                             ),
250
-                            tabPanel("QC",
251
-                                     br(),
252
-                                     p(paste0("This panel shows the absolute",
265
+                            shiny::tabPanel("QC",
266
+                                     shiny::br(),
267
+                                     shiny::p(paste0("This panel shows the absolute",
253 268
                                               " correlations between",
254 269
                                               " Principal",
255 270
                                               " Components (PCs) of",
... ...
@@ -264,30 +279,30 @@ sconeReport = function(x, methods,
264 279
                                               " could have a low value of ",
265 280
                                               "PC1 and 'bad' cells a",
266 281
                                               " high value).")),
267
-                                     br(),
282
+                                     shiny::br(),
268 283
                                      plotlyOutput('qccorPlot',
269 284
                                                   width = "650px",
270 285
                                                   height = "450px"),
271
-                                     h5("PCA of QC metrics"),
272
-                                     h6("2D"),
273
-                                     plotOutput("plot_qc",
286
+                                     shiny::h5("PCA of QC metrics"),
287
+                                     shiny::h6("2D"),
288
+                                     shiny::plotOutput("plot_qc",
274 289
                                                 width = "650px",
275 290
                                                 height = "450px"),
276
-                                     h6("3D Interactive"),
291
+                                     shiny::h6("3D Interactive"),
277 292
                                      plotlyOutput("plot3d_qc",
278 293
                                                   width = "650px",
279 294
                                                   height = "650px")
280 295
                             ),
281
-                            tabPanel("Silhouette",
282
-                                     br(),
283
-                                     p(paste0("Silhouette Width per",
296
+                            shiny::tabPanel("Silhouette",
297
+                                     shiny::br(),
298
+                                     shiny::p(paste0("Silhouette Width per",
284 299
                                               " Sample and Contingency",
285 300
                                               " Tables")),
286
-                                     br(),
287
-                                     plotOutput("plotsil",
301
+                                     shiny::br(),
302
+                                     shiny::plotOutput("plotsil",
288 303
                                                 width = "650px",
289 304
                                                 height = "450px"),
290
-                                     selectInput("cat1", 
305
+                                     shiny::selectInput("cat1",
291 306
                                                  label =
292 307
                                                    "Row Class",
293 308
                                                  choices =
... ...
@@ -297,7 +312,7 @@ sconeReport = function(x, methods,
297 312
                                                           "bio", "PAM Cluster"=
298 313
                                                           "clust"),
299 314
                                                  selected = "batch"),
300
-                                     selectInput("cat2",
315
+                                     shiny::selectInput("cat2",
301 316
                                                  label = "Column Class",
302 317
                                                  choices =
303 318
                                                    list("Batch"=
... ...
@@ -307,33 +322,33 @@ sconeReport = function(x, methods,
307 322
                                                         "PAM Cluster"=
308 323
                                                           "clust"),
309 324
                                                  selected = "bio"),
310
-                                     tableOutput("cat_tab")
325
+                                     shiny::tableOutput("cat_tab")
311 326
                             ),
312
-                            tabPanel("Control Genes",
313
-                                     br(),
314
-                                     p(paste0("Heatmap of control genes, ",
327
+                            shiny::tabPanel("Control Genes",
328
+                                     shiny::br(),
329
+                                     shiny::p(paste0("Heatmap of control genes, ",
315 330
                                               "colored by all",
316 331
                                               " three categories.")),
317
-                                     br(),
318
-                                     p("Positive controls:"),
319
-                                     plotOutput("hmposcon",
332
+                                     shiny::br(),
333
+                                     shiny::p("Positive controls:"),
334
+                                     shiny::plotOutput("hmposcon",
320 335
                                                 width = "650px",
321 336
                                                 height = "450px"),
322
-                                     p("Negative controls:"),
323
-                                     plotOutput("hmnegcon",
337
+                                     shiny::p("Negative controls:"),
338
+                                     shiny::plotOutput("hmnegcon",
324 339
                                                 width = "650px",
325 340
                                                 height = "450px")
326 341
                             ),
327
-                            
328
-                            tabPanel("Stratified PCA",
329
-                                     br(),
330
-                                     p(paste0("Violin plots showing",
342
+
343
+                            shiny::tabPanel("Stratified PCA",
344
+                                     shiny::br(),
345
+                                     shiny::p(paste0("Violin plots showing",
331 346
                                               " conditional ",
332 347
                                               "distributions of",
333 348
                                               " selected Principal",
334 349
                                               " Component.")),
335
-                                     br(),
336
-                                     sliderInput("pcsel", 
350
+                                     shiny::br(),
351
+                                     shiny::sliderInput("pcsel",
337 352
                                                  label =
338 353
                                                    paste0("Selected",
339 354
                                                           " Principal ",
... ...
@@ -341,11 +356,11 @@ sconeReport = function(x, methods,
341 356
                                                  min=1, max=
342 357
                                                    min(length(batch)-1,10),
343 358
                                                  value = 1),
344
-                                     h6("All Genes:"),
345
-                                     plotOutput("violin_base",
359
+                                     shiny::h6("All Genes:"),
360
+                                     shiny::plotOutput("violin_base",
346 361
                                                 width = "650px",
347 362
                                                 height = "450px"),
348
-                                     selectInput("gene_set2", 
363
+                                     shiny::selectInput("gene_set2",
349 364
                                                  label = "Gene selection",
350 365
                                                  choices =
351 366
                                                    list("Negative Controls"=
... ...
@@ -353,74 +368,74 @@ sconeReport = function(x, methods,
353 368
                                                         "Positive Controls"=
354 369
                                                           "pos"),
355 370
                                                  selected = "neg"),
356
-                                     h6("Select Genes:"),
357
-                                     plotOutput("violin_select",
371
+                                     shiny::h6("Select Genes:"),
372
+                                     shiny::plotOutput("violin_select",
358 373
                                                 width = "650px",
359 374
                                                 height = "450px")
360 375
                             ),
361
-                            
362
-                            tabPanel("Relative Log-Expression",
363
-                                     br(),
364
-                                     p(paste0("Relative ",
376
+
377
+                            shiny::tabPanel("Relative Log-Expression",
378
+                                     shiny::br(),
379
+                                     shiny::p(paste0("Relative ",
365 380
                                               "Log-Expression Plot ",
366 381
                                               "for top 100 Most ",
367 382
                                               "Variable Genes.")),
368
-                                     plotOutput("rle",
383
+                                     shiny::plotOutput("rle",
369 384
                                                 width = "850px",
370 385
                                                 height = "450px")
371 386
                             )
372 387
       ))))
373
-  
388
+
374 389
   server <- function(input, output, session) {
375
-    
390
+
376 391
     ## ------ Overview Tab ------
377
-    
392
+
378 393
     # Render Network
379 394
     output$norm_net <- renderVisNetwork({
380
-      
395
+
381 396
       # Awk: Check if any non-loaded methods are included
382 397
       if(any(!all_nodes  %in% leaves)){
383
-        visNetwork(nodes, edges,width = "100%",main = "Tree of Methods") %>% 
384
-          visHierarchicalLayout(sortMethod = "directed" )  %>% 
385
-          visGroups(groupname = "NoData", shape = "dot", 
398
+        visNetwork(nodes, edges,width = "100%",main = "Tree of Methods") %>%
399
+          visHierarchicalLayout(sortMethod = "directed" )  %>%
400
+          visGroups(groupname = "NoData", shape = "dot",
386 401
                     size = 10, color = list(background = "lightgrey",
387 402
                                             border = "darkblue",
388
-                                            highlight = 
403
+                                            highlight =
389 404
                                               list(background =
390 405
                                                      "black",
391
-                                                   border = "red")), 
392
-                    shadow = list(enabled = TRUE)) %>% 
393
-          visGroups(groupname = "Loaded", shape = "dot", 
406
+                                                   border = "red")),
407
+                    shadow = list(enabled = TRUE)) %>%
408
+          visGroups(groupname = "Loaded", shape = "dot",
394 409
                     size = 20, color = list(background = "lightblue",
395
-                                            border = "darkblue", 
396
-                                            highlight = 
410
+                                            border = "darkblue",
411
+                                            highlight =
397 412
                                               list(background = "cyan",
398 413
                                                    border = "red")),
399 414
                     shadow = list(enabled = TRUE)) %>%
400 415
           visEdges(shadow = TRUE, arrows = list(to = list(enabled = TRUE)),
401 416
                    color = list(color = "darkblue", highlight = "red")) %>%
402 417
           visOptions(nodesIdSelection =
403
-                       list(enabled = TRUE, 
418
+                       list(enabled = TRUE,
404 419
                             values = nodes$id[nodes$group == "Loaded"],
405 420
                             selected =
406 421
                               nodes$id[nodes$title ==
407 422
                                          rownames(scone_res$params)[1]])) %>%
408 423
           visLegend(width = 0.1, position = "right", main = "Status")
409 424
       }else{
410
-        visNetwork(nodes, edges,width = "100%",main = "Tree of Methods") %>% 
411
-          visHierarchicalLayout(sortMethod = "directed" )  %>% 
425
+        visNetwork(nodes, edges,width = "100%",main = "Tree of Methods") %>%
426
+          visHierarchicalLayout(sortMethod = "directed" )  %>%
412 427
           visEdges(shadow = TRUE, arrows = list(to = list(enabled = TRUE)),
413 428
                    color = list(color = "darkblue", highlight = "red")) %>%
414 429
           visOptions(nodesIdSelection =
415
-                       list(enabled = TRUE, 
430
+                       list(enabled = TRUE,
416 431
                             values = nodes$id[nodes$group == "Loaded"],
417
-                            selected = 
432
+                            selected =
418 433
                               nodes$id[nodes$title ==
419 434
                                          rownames(scone_res$params)[1]]))
420 435
       }
421
-      
436
+
422 437
     })
423
-    
438
+
424 439
     # Render Table
425 440
     output$tbl_score <- DT::renderDataTable({
426 441
       DT::datatable(scone_res$scores,
... ...
@@ -432,21 +447,21 @@ sconeReport = function(x, methods,
432 447
                       list(columnDefs =
433 448
                              list(list(visible=FALSE,
434 449
                                        targets=1:(ncol(scone_res$scores)-1))),
435
-                           dom = 'Bfrtip', 
450
+                           dom = 'Bfrtip',
436 451
                            buttons = list(
437 452
                              list(extend = 'colvis',
438 453
                                   columns = c(1:(ncol(scone_res$scores)-1))))
439 454
                       )) %>%
440 455
         DT::formatSignif(columns=c(1:(ncol(scone_res$scores))), digits=3)
441 456
     })
442
-    
457
+
443 458
     # Update Menu Upon Network Selection of (Loaded) Normalization
444
-    observeEvent(input$norm_net_selected,{
445
-      if(is.character(input$norm_net_selected)){ # probably unnecessary 
459
+    shiny::observeEvent(input$norm_net_selected,{
460
+      if(is.character(input$norm_net_selected)){ # probably unnecessary
446 461
         if(!input$norm_net_selected == ""){ # probably unnecessary
447 462
           if(as.integer(input$norm_net_selected) %in%
448 463
              nodes$id[nodes$group == "Loaded"]){
449
-            updateSelectInput(
464
+            shiny::updateSelectInput(
450 465
               session,
451 466
               "norm_code",
452 467
               selected = as.character(
... ...
@@ -457,12 +472,12 @@ sconeReport = function(x, methods,
457 472
         }
458 473
       }
459 474
     })
460
-    
475
+
461 476
     # Update Menu Upon Table Selection
462
-    observeEvent(input$tbl_score_rows_selected,{
463
-      if(length(input$tbl_score_rows_selected) > 0 ){ # probably unnecessary 
464
-        updateSelectInput(session,
465
-                          "norm_code", 
477
+    shiny::observeEvent(input$tbl_score_rows_selected,{
478
+      if(length(input$tbl_score_rows_selected) > 0 ){ # probably unnecessary
479
+        shiny::updateSelectInput(session,
480
+                          "norm_code",
466 481
                           selected = as.character(
467 482
                             nodes$title[nodes$id ==
468 483
                                           as.integer(
... ...
@@ -472,22 +487,22 @@ sconeReport = function(x, methods,
472 487
         )
473 488
       }
474 489
     })
475
-    
490
+
476 491
     # Upon Menu Selection of Normalization
477
-    observeEvent(input$norm_code,{
478
-      
479
-      # Update Network 
492
+    shiny::observeEvent(input$norm_code,{
493
+
494
+      # Update Network
480 495
       if(is.character(input$norm_net_selected)){
481
-        
496
+
482 497
         # If selection is empty, then no valid node is selected.
483 498
         if(input$norm_net_selected == ""){
484 499
           visNetworkProxy("norm_net") %>%
485 500
             visSelectNodes(id =
486 501
                              array(nodes$id[nodes$title == input$norm_code]))
487 502
         }else{
488
-          
503
+
489 504
           # If selection is different from menu, then network must be updated.
490
-          if(as.character(nodes$title[nodes$id == 
505
+          if(as.character(nodes$title[nodes$id ==
491 506
                                       as.integer(
492 507
                                         input$norm_net_selected
493 508
                                         )]) != input$norm_code){
... ...
@@ -497,120 +512,120 @@ sconeReport = function(x, methods,
497 512
           }
498 513
         }
499 514
       }
500
-      
515
+
501 516
       # Update Table
502
-      
517
+
503 518
       # If selection is empty, then no row is selected.
504 519
       if(length(input$tbl_score_rows_selected) == 0){
505
-        dataTableProxy("tbl_score") %>%
506
-          selectRows(nodes$id[nodes$title == input$norm_code])
520
+        DT::dataTableProxy("tbl_score") %>%
521
+          DT::selectRows(nodes$id[nodes$title == input$norm_code])
507 522
       }else{
508
-        
523
+
509 524
         # If selection is different from menu, then row must be updated.
510 525
         if(input$tbl_score_rows_selected !=
511 526
            nodes$id[nodes$title == input$norm_code]){
512
-          dataTableProxy("tbl_score") %>%
513
-            selectRows(nodes$id[nodes$title == input$norm_code])
527
+          DT::dataTableProxy("tbl_score") %>%
528
+            DT::selectRows(nodes$id[nodes$title == input$norm_code])
514 529
         }
515 530
       }
516
-      
531
+
517 532
     })
518
-    
533
+
519 534
     ## ------ PCA Tab ------
520
-    
521
-    normle <- reactive({
535
+
536
+    normle <- shiny::reactive({
522 537
       as.matrix(scone_res$normalized_data[[input$norm_code]])
523 538
     })
524
-    
525
-    strat_col <- reactive({
539
+
540
+    strat_col <- shiny::reactive({
526 541
       switch(input$color_code,
527 542
              bio = bio,
528 543
              batch = batch,
529 544
              clust = pam_obj()$clust)
530 545
     })
531
-    
532
-    pc_col <- reactive({
546
+
547
+    pc_col <- shiny::reactive({
533 548
       cc[strat_col()]
534 549
     })
535
-    
536
-    pc_gene <- reactive({
550
+
551
+    pc_gene <- shiny::reactive({
537 552
       switch(input$gene_set,
538 553
              pos = intersect(poscon,rownames(normle())),
539 554
              neg = intersect(negcon,rownames(normle())))
540 555
     })
541
-    
542
-    pc_obj_base <- reactive({
543
-      
556
+
557
+    pc_obj_base <- shiny::reactive({
558
+
544 559
       # If user has not provided evaluation projection
545 560
       if(is.null(eval_proj)){
546
-        
561
+
547 562
         prcomp(t(normle()),
548 563
                center = TRUE, scale = TRUE)
549
-        
564
+
550 565
       } else {
551
-        
566
+
552 567
         proj = eval_proj(normle(),
553 568
                          eval_proj_args = eval_proj_args)
554 569
         colnames(proj) = paste0("PC",1:ncol(proj))
555
-        
570
+
556 571
         pc_out = list(x = proj, sdev = apply(proj,2,sd))
557
-        
572
+
558 573
         pc_out
559
-        
574
+
560 575
       }
561
-      
576
+
562 577
     })
563
-    
564
-    pc_obj_select <- reactive({
578
+
579
+    pc_obj_select <- shiny::reactive({
565 580
       if(length(pc_gene()) > 0){
566
-        
581
+
567 582
         # If user has not provided evaluation projection
568 583
         if(is.null(eval_proj)){
569
-          
584
+
570 585
           prcomp(t(normle()[pc_gene(),]),
571 586
                  center = TRUE, scale = TRUE)
572
-          
587
+
573 588
         } else {
574
-          
589
+
575 590
           proj = eval_proj(normle()[pc_gene(),],
576 591
                            eval_proj_args = eval_proj_args)
577 592
           colnames(proj) = paste0("PC",1:ncol(proj))
578
-          
593
+
579 594
           pc_out = list(x = proj, sdev = apply(proj,2,sd))
580
-          
595
+
581 596
           pc_out
582
-          
597
+
583 598
         }
584
-        
599
+
585 600
       }else{
586 601
         list()
587 602
       }
588 603
     })
589
-    
590
-    pam_obj <- reactive({
604
+
605
+    pam_obj <- shiny::reactive({
591 606
       cluster::pam(x = pc_obj_base()$x[,1:input$dim],k = input$k)
592 607
     })
593
-    
594
-    output$plot_scree <- renderPlot({
608
+
609
+    output$plot_scree <- shiny::renderPlot({
595 610
       plot(pc_obj_base()$sdev[1:min(input$dim*2,
596 611
                                     length(batch)-1)]^2/
597
-             sum(pc_obj_base()$sdev^2), 
612
+             sum(pc_obj_base()$sdev^2),
598 613
            typ = "l",
599
-           ylab = "Proportion of Variance", 
614
+           ylab = "Proportion of Variance",
600 615
            xlab = "PC Index",
601 616
            log = "y")
602 617
       abline(v = input$dim, col = "red",lty = 2)
603 618
     })
604
-    
605
-    output$plot_base <- renderPlot({
619
+
620
+    output$plot_base <- shiny::renderPlot({
606 621
       par(mar=c(5.1, 4.1, 4.1, 10.1), xpd=TRUE)
607 622
       plot(pc_obj_base()$x[,1:2],col =  pc_col(), pch = 16)
608
-      legend("topright", inset=c(-0.3,0), 
609
-             legend=levels(factor(strat_col())), 
623
+      legend("topright", inset=c(-0.3,0),
624
+             legend=levels(factor(strat_col())),
610 625
              fill = cc[sort(unique(factor(strat_col())))])
611 626
     })
612
-    
613
-    output$plot3d_base <- renderPlotly({
627
+
628
+    output$plot3d_base <- shiny::renderPlotly({
614 629
       PC1 <- PC2 <- PC3 <- NULL
615 630
       df <- setNames(data.frame(pc_obj_base()$x[,1:3]),
616 631
                      c("PC1", "PC2", "PC3"))
... ...
@@ -618,96 +633,96 @@ sconeReport = function(x, methods,
618 633
               type = "scatter3d", mode = "markers",
619 634
               marker = list(color=pc_col() ))
620 635
     })
621
-    
622
-    output$plot_select <- renderPlot({
623
-      
636
+
637
+    output$plot_select <- shiny::renderPlot({
638
+
624 639
       par(mar=c(5.1, 4.1, 4.1, 10.1), xpd=TRUE)
625
-      
640
+
626 641
       if(length(pc_gene()) > 0){
627
-        
642
+
628 643
         plot(pc_obj_select()$x[,1:2],
629 644
              col =  pc_col(), pch = 16)
630
-        legend("topright", inset=c(-0.3,0), 
645
+        legend("topright", inset=c(-0.3,0),
631 646
                legend=levels(factor(strat_col())),
632 647
                fill = cc[sort(unique(factor(strat_col())))])
633
-        
648
+
634 649
       }else{
635
-        
650
+
636 651
         plot(0, type = 'n',xlab = "",ylab = "",xaxt = 'n',yaxt = 'n')
637 652
         text(0,labels = "Control gene set is empty.")
638
-        
653
+
639 654
       }
640
-      
655
+
641 656
     })
642
-    
643
-    pam_obj <- reactive({
657
+
658
+    pam_obj <- shiny::reactive({
644 659
       cluster::pam(x = pc_obj_base()$x[,1:input$dim],k = input$k)
645 660
     })
646
-    
661
+
647 662
     ## ------ QC Tab ------
648
-    
649
-    cor_qc <- reactive({
663
+
664
+    cor_qc <- shiny::reactive({
650 665
       abs(cor(qc,pc_obj_base()$x))
651 666
     })
652
-    
653
-    output$qccorPlot <- renderPlotly({
667
+
668
+    output$qccorPlot <- shiny::renderPlotly({
654 669
       metric <- value <- PC <- NULL
655
-      
656
-      df = melt(cor_qc()[,1:input$dim])
670
+
671
+      df = reshape2::melt(cor_qc()[,1:input$dim])
657 672
       colnames(df) = c("metric","PC","value")
658 673
       df$metric = factor(df$metric,levels = colnames(qc))
659 674
       df$PC = factor(df$PC,
660 675
                      levels = paste0("PC",1:input$dim))
661
-      p <- ggplot(data = df, aes(x = PC, 
662
-                                 fill = metric, 
676
+      p <- ggplot(data = df, aes(x = PC,
677
+                                 fill = metric,
663 678
                                  weight=value)) +
664
-        geom_bar(position = "dodge") + 
665
-        ylim(0, 1) + 
679
+        geom_bar(position = "dodge") +
680
+        ylim(0, 1) +
666 681
         labs(x="PC of Expression",
667
-             y="Absolute Correlation") + 
682
+             y="Absolute Correlation") +
668 683
         theme(legend.title=element_blank())
669 684
       ggplotly(p)
670 685
     })
671
-    
672
-    pc_obj_qc <- reactive({
686
+
687
+    pc_obj_qc <- shiny::reactive({
673 688
       prcomp(as.matrix(qc),center = TRUE, scale = TRUE)
674 689
     })
675
-    
676
-    output$plot_qc <- renderPlot({
690
+
691
+    output$plot_qc <- shiny::renderPlot({
677 692
       par(mar=c(5.1, 4.1, 4.1, 10.1), xpd=TRUE)
678 693
       plot(pc_obj_qc()$x[,1:2],col =  pc_col(), pch = 16)
679 694
       legend("topright", inset=c(-0.3,0),
680 695
              legend=levels(factor(strat_col())),
681 696
              fill = cc[sort(unique(factor(strat_col())))])
682 697
     })
683
-    
684
-    output$plot3d_qc <- renderPlotly({
698
+
699
+    output$plot3d_qc <- shiny::renderPlotly({
685 700
       if(ncol(pc_obj_qc()$x) >= 3){
686 701
         PC1 <- PC2 <- PC3 <- NULL
687 702
         df <- setNames(data.frame(pc_obj_qc()$x[,1:3]),
688 703
                        c("PC1", "PC2", "PC3"))
689
-        plot_ly(df, x = ~PC1, y = ~PC2, z = ~PC3, 
690
-                type = "scatter3d", 
704
+        plot_ly(df, x = ~PC1, y = ~PC2, z = ~PC3,
705
+                type = "scatter3d",
691 706
                 mode = "markers",
692 707
                 marker = list(color=pc_col() ))
693 708
       }else{
694 709
         plot_ly(data.frame(), type = "scatter3d", mode = "markers")
695 710
       }
696 711
     })
697
-    
712
+
698 713
     ## ------ Silhouette Tab ------
699
-    
700
-    sil_obj <- reactive({
714
+
715
+    sil_obj <- shiny::reactive({
701 716
       cluster::silhouette(x = as.numeric(strat_col()),
702 717
                           dist = dist(pc_obj_base()$x[,1:input$dim]))
703 718
     })
704
-    
705
-    output$plotsil <- renderPlot({
706
-      
719
+
720
+    output$plotsil <- shiny::renderPlot({
721
+
707 722
       par(mar=c(5.1, 4.1, 4.1, 10.1), xpd=TRUE)
708
-      
723
+
709 724
       if(length(unique(strat_col())) > 1){
710
-        
725
+
711 726
         sil = sil_obj()
712 727
         o1 = order(sil[,3])
713 728
         o = o1[order(-sil[,1][o1])]
... ...
@@ -717,205 +732,205 @@ sconeReport = function(x, methods,
717 732
                 xlim = 1.5*range(c(sil[,3],-sil[,3])),
718 733
                 col = pc_col()[o],
719 734
                 border = pc_col()[o])
720
-        legend("topright", inset=c(-0.3,0), 
721
-               legend=levels(factor(strat_col())), 
735
+        legend("topright", inset=c(-0.3,0),
736
+               legend=levels(factor(strat_col())),
722 737
                fill = cc[sort(unique(factor(strat_col())))])
723
-        
738
+
724 739
       }else{
725
-        
740
+
726 741
         plot(0, type = 'n',xlab = "",ylab = "",xaxt = 'n',yaxt = 'n')
727 742
         text(0,labels = "Stratify plots by a multi-level classification.")
728
-        
743
+
729 744
       }
730
-      
745
+
731 746
     })
732
-    
733
-    cat1 <- reactive({
747
+
748
+    cat1 <- shiny::reactive({
734 749
       switch(input$cat1,
735 750
              bio = bio,
736 751
              batch = batch,
737 752
              clust = pam_obj()$clust)
738 753
     })
739
-    
740
-    cat2 <- reactive({
754
+
755
+    cat2 <- shiny::reactive({
741 756
       switch(input$cat2,
742 757
              bio = bio,
743 758
              batch = batch,
744 759
              clust = pam_obj()$clust)
745 760
     })
746
-    
747
-    output$cat_tab <- renderTable({
748
-      table(cat1(), cat2())      
761
+
762
+    output$cat_tab <- shiny::renderTable({
763
+      table(cat1(), cat2())
749 764
     })
750
-    
751
-    
765
+
766
+
752 767
     ## ------ Control Genes Tab ------
753
-    
754
-    silo <- reactive({
768
+
769
+    silo <- shiny::reactive({
755 770
       sil = sil_obj()
756 771
       o1 = order(sil[,3])
757
-      o1[order(-sil[,1][o1])]    
772
+      o1[order(-sil[,1][o1])]
758 773
     })
759
-    
760
-    output$hmnegcon <- renderPlot({
761
-      
774
+
775
+    output$hmnegcon <- shiny::renderPlot({
776
+
762 777
       if(length(negcon) > 0){
763
-        
778
+
764 779
         if(length(unique(strat_col())) > 1){
765
-          
766
-          aheatmap(normle()[negcon,], Colv = silo(), 
780
+
781
+          NMF::aheatmap(normle()[negcon,], Colv = silo(),
767 782
                    annCol = list(batch = batch, bio = bio,
768 783
                                  pam =  as.factor(pam_obj()$clust)),
769 784
                    annColors = list(batch = cc, bio = cc, pam =  cc))
770
-          
785
+
771 786
         }else{
772
-          
787
+
773 788
           plot(0, type = 'n',xlab = "",ylab = "",xaxt = 'n',yaxt = 'n')
774 789
           text(0,labels = "Stratify plots by a multi-level classification.")
775
-          
790
+
776 791
         }
777
-        
792
+
778 793
       }else{
779
-        
794
+
780 795
         plot(0, type = 'n',xlab = "",ylab = "",xaxt = 'n',yaxt = 'n')
781 796
         text(0,labels = "Control gene set is empty.")
782
-        
797
+
783 798
       }
784
-      
799
+
785 800
     })
786
-    
787
-    output$hmposcon <- renderPlot({
788
-      
801
+
802
+    output$hmposcon <- shiny::renderPlot({
803
+
789 804
       if(length(poscon) > 0){
790
-        
805
+
791 806
         if(length(unique(strat_col())) > 1){
792
-          
793
-          aheatmap(normle()[poscon,], Colv = silo(), 
794
-                   annCol = list(batch = batch, 
795
-                                 bio = bio, 
807
+
808
+          NMF::aheatmap(normle()[poscon,], Colv = silo(),
809
+                   annCol = list(batch = batch,
810
+                                 bio = bio,
796 811
                                  pam = as.factor(pam_obj()$clust)),
797 812
                    annColors = list(batch = cc, bio = cc, pam =  cc))
798
-          
813
+
799 814
         }else{
800
-          
815
+
801 816
           plot(0, type = 'n',xlab = "",ylab = "",xaxt = 'n',yaxt = 'n')
802 817
           text(0,labels = "Stratify plots by a multi-level classification.")
803
-          
818
+
804 819
         }
805
-        
820
+
806 821
       }else{
807
-        
822
+
808 823
         plot(0, type = 'n',xlab = "",ylab = "",xaxt = 'n',yaxt = 'n')
809 824
         text(0,labels = "Control gene set is empty.")
810
-        
825
+
811 826
       }
812
-      
827
+
813 828
     })
814
-    
829
+
815 830
     ## ------ Stratified PCA Tab ------
816
-    
817
-    output$violin_base <- renderPlot({
818
-      
831
+
832
+    output$violin_base <- shiny::renderPlot({
833
+
819 834
       Class = factor(strat_col())
820
-      Val = pc_obj_base()$x[,input$pcsel] 
821
-      
822
-      ggplot(data.frame(Class,Val ),aes(x = Class,y = Val))   +  
823
-        geom_violin(scale = "width", trim = TRUE, aes(fill = Class))+ 
835
+      Val = pc_obj_base()$x[,input$pcsel]
836
+
837
+      ggplot(data.frame(Class,Val ),aes(x = Class,y = Val))   +
838
+        geom_violin(scale = "width", trim = TRUE, aes(fill = Class))+
824 839
         labs(x = "Plot Stratification", y = "PC") +
825 840
         coord_cartesian(ylim = max(abs(range(Val)))*c(-1.5,1.5))  +
826 841
         scale_fill_manual(values=cc[sort(unique(factor(strat_col())))]) +
827 842
         geom_point(colour = "black") + guides(fill=FALSE)
828
-      
843
+
829 844
     })
830
-    
831
-    pc_gene2 <- reactive({
845
+
846
+    pc_gene2 <- shiny::reactive({
832 847
       switch(input$gene_set2,
833 848
              pos = intersect(poscon,rownames(normle())),
834 849
              neg = intersect(negcon,rownames(normle())))
835 850
     })
836
-    
837
-    pc_obj_select2 <- reactive({
838
-      
851
+
852
+    pc_obj_select2 <- shiny::reactive({
853
+
839 854
       # If selected genes exist
840 855
       if(length(pc_gene2()) > 0){
841
-        
856
+
842 857
         # If user has not provided evaluation projection
843 858
         if(is.null(eval_proj)){
844
-          
859
+
845 860
           prcomp(t(normle()[pc_gene2(),]),
846 861
                  center = TRUE, scale = TRUE)
847
-          
862
+
848 863
         } else {
849
-          
864
+
850 865
           proj = eval_proj(normle()[pc_gene2(),],
851 866
                            eval_proj_args = eval_proj_args)
852 867
           colnames(proj) = paste0("PC",1:ncol(proj))
853
-          
868
+
854 869
           pc_out = list(x = proj, sdev = apply(proj,2,sd))
855
-          
870
+
856 871
           pc_out
857
-          
872
+
858 873
         }
859
-        
874
+
860 875
       }else{
861 876
         list()
862 877
       }
863
-      
878
+
864 879
     })
865
-    
866
-    output$violin_select <- renderPlot({
867
-      
880
+
881
+    output$violin_select <- shiny::renderPlot({
882
+
868 883
       if(length(pc_gene2()) > 0){
869
-        
884
+
870 885
         Class = factor(strat_col())
871
-        Val = pc_obj_select2()$x[,input$pcsel] 
872
-        
873
-        ggplot(data.frame(Class,Val ),aes(x = Class,y = Val))   +  
874
-          geom_violin(scale = "width", trim = TRUE, aes(fill = Class))+ 
886
+        Val = pc_obj_select2()$x[,input$pcsel]
887
+
888
+        ggplot(data.frame(Class,Val ),aes(x = Class,y = Val))   +
889
+          geom_violin(scale = "width", trim = TRUE, aes(fill = Class))+
875 890
           labs(x = "Plot Stratification", y = "PC") +
876 891
           coord_cartesian(ylim = max(abs(range(Val)))*c(-1.5,1.5))  +
877 892
           scale_fill_manual(values=cc[sort(unique(factor(strat_col())))]) +
878 893
           geom_point(colour = "black") + guides(fill=FALSE)
879
-        
894
+
880 895
       }else{
881
-        
896
+
882 897
         plot(0, type = 'n',xlab = "",ylab = "",xaxt = 'n',yaxt = 'n')
883 898
         text(0,labels = "Control gene set is empty.")
884
-        
899
+
885 900
       }
886 901
     })
887
-    
902
+
888 903
     ## ------ Relative Log-Expression ------
889
-    
890
-    output$rle <- renderPlot({
891
-      
904
+
905
+    output$rle <- shiny::renderPlot({
906
+
892 907
       par(mar=c(5.1, 4.1, 4.1, 10.1), xpd=TRUE)
893
-      
908
+
894 909
       if(length(unique(strat_col())) > 1){
895
-        
910
+
896 911
         vars = apply(exp(normle()),1,var)
897 912
         is_var = rank(-vars) <= 100
898 913
         median <- apply(normle()[is_var,], 1, median)
899 914
         rle <- apply(normle()[is_var,], 2, function(x) x - median)
900
-        boxplot(rle[,rev(silo())],col = pc_col()[rev(silo())], 
915
+        boxplot(rle[,rev(silo())],col = pc_col()[rev(silo())],
901 916
                 outline = FALSE, names = rep("",ncol(rle)))
902 917
         abline(h=0, lty=2)
903 918
         legend("topright", inset=c(-0.2,0),
904
-               legend=levels(factor(strat_col())), 
919
+               legend=levels(factor(strat_col())),
905 920
                fill = cc[sort(unique(factor(strat_col())))])
906
-        
921
+
907 922
       }else{
908
-        
923
+
909 924
         plot(0, type = 'n',xlab = "",ylab = "",xaxt = 'n',yaxt = 'n')
910 925
         text(0,labels = "Stratify plots by a multi-level classification.")
911
-        
926
+
912 927
       }
913
-      
928
+
914 929
     })
915
-    
930
+
916 931
     ## ----- Download Button -----
917
-    
918
-    output$downloadData <- downloadHandler(
932
+
933
+    output$downloadData <- shiny::downloadHandler(
919 934
       filename = function() {
920 935
         paste('scone_out-', Sys.Date(), '.csv', sep='')
921 936
       },
... ...
@@ -929,10 +944,10 @@ sconeReport = function(x, methods,
929 944
         write.csv(datt, con)
930 945
       }
931 946
     )
932
-    
947
+
933 948
   }
934
-  
949
+
935 950
   # Shiny App
936
-  shinyApp(ui = ui, server = server)
937
-  
951
+  shiny::shinyApp(ui = ui, server = server)
952
+
938 953
 }
Browse code

Merge with bioc submission

mbcole authored on 10/01/2017 00:14:56
Showing0 changed files
Browse code

Modified documentation, more examples and exported getters, added links between SconeExperiment and methods, and apply getters to setValidate

mbcole authored on 14/12/2016 00:54:28
Showing1 changed files
... ...
@@ -38,7 +38,7 @@
38 38
 #' set.seed(101)
39 39
 #' mat <- matrix(rpois(1000, lambda = 5), ncol=10)
40 40
 #' colnames(mat) <- paste("X", 1:ncol(mat), sep="")
41
-#' obj <- sconeExperiment(mat)
41
+#' obj <- SconeExperiment(mat)
42 42
 #' res <- scone(obj, scaling=list(none=identity, uq=UQ_FN, deseq=DESEQ_FN),
43 43
 #'            evaluate=TRUE, k_ruv=0, k_qc=0, eval_kclust=2, 
44 44
 #'            bpparam = BiocParallel::SerialParam())
Browse code

Added custom projection to report

mbcole authored on 03/12/2016 21:47:58
Showing1 changed files
... ...
@@ -16,6 +16,11 @@
16 16
 #' @param poscon character. Genes to be used as positive controls for 
17 17
 #'   evaluation. These genes should be expected to change according to the 
18 18
 #'   biological phenomenon of interest. Default empty character.
19
+#' @param eval_proj function. Projection function for evaluation  (see 
20
+#'   \code{\link{score_matrix}} for details). If NULL, PCA is used for 
21
+#'   projection.
22
+#' @param eval_proj_args list. List of args passed to projection function as
23
+#'   eval_proj_args.
19 24
 #'   
20 25
 #' @importFrom shiny fluidPage downloadLink br titlePanel tableOutput helpText
21 26
 #'   downloadHandler column h5 h6 updateSelectInput plotOutput tabsetPanel
... ...
@@ -30,6 +35,7 @@
30 35
 #' @importFrom NMF aheatmap
31 36
 #' @importFrom ggplot2 labs theme geom_point ylim ggplot geom_violin
32 37
 #'   element_blank aes geom_bar coord_cartesian scale_fill_manual guides
38
+#' @importFrom rARPACK svds
33 39
 #' @export
34 40
 #' 
35 41
 #' @return An object that represents the SCONE report app.
... ...
@@ -52,7 +58,9 @@
52 58
 sconeReport = function(x, methods,
53 59
                        qc, 
54 60
                        bio = NULL, batch = NULL,
55
-                       poscon = character(), negcon = character()){
61
+                       poscon = character(), negcon = character(),
62
+                       eval_proj = NULL,
63
+                       eval_proj_args = NULL){
56 64
   
57 65
   
58 66
   scone_res = list()
... ...
@@ -532,12 +540,48 @@ sconeReport = function(x, methods,
532 540
     })
533 541
     
534 542
     pc_obj_base <- reactive({
535
-      prcomp(t(normle()),center = TRUE, scale = TRUE)
543
+