Improved Plots

.gitignore

@@ -7,4 +7,5 @@ extdata/
 inst/doc
 /doc/
 /Meta/
-.Rbuildignore
+.Rbuildignore
+test_dev

DESCRIPTION

@@ -1,7 +1,7 @@
 Package: scMaSigPro
 Type: Package
 Title: Application of MaSigPro Bioconductor Package for scRNA Trajectory data
-Version: 0.0.1
+Version: 0.0.2
 Authors@R: c(
     person("Ana", "Conesa", role = c("aut"), email = "ana.conesa@csic.es"),
     person("Maria Jose", "Nueda", role = c("aut"), email = "mj.nueda@ua.es"),

R/plotTrend.R

@@ -41,6 +41,9 @@ plotTrend <- function(scmpObj,
   pooled.time <- "pooled.time"
   path <- "path"
 
+  # Offset
+  offset_vector <- scmpObj@Design@offset
+
   # Check summary_mode
   assert_that(any(summary_mode %in% c("median", "mean")),
     msg = paste(
@@ -144,6 +147,11 @@ plotTrend <- function(scmpObj,
   data.sol <- showSol(scmpObj, view = FALSE, return = TRUE)
   data.sol <- data.sol[feature_id, , drop = FALSE]
 
+  # Correct by offset
+  if (sum(offset_vector) != 0) {
+    points.df["pb.counts"] <- points.df["pb.counts"] / exp(offset_vector)
+  }
+
   # if log is requestion
   if (logs) {
     if (logType == "log2") {
@@ -174,10 +182,14 @@ plotTrend <- function(scmpObj,
       summarize(pb.counts = median(pb.counts), .groups = "drop")
   }
 
+  if (sum(offset_vector) != 0) {
+    line.df <- points.df
+  }
+
   # Plot
   p <- ggplot() +
     geom_point(data = points.df, aes(x = pooled.time, y = pb.counts, color = path), fill = "#102C57", alpha = 0.5, size = 2, stroke = 1, shape = 21) +
-    geom_line(data = line.df, aes(x = pooled.time, y = .data$pb.counts, color = path), linetype = "solid", linewidth = 1, alpha = 0.7) +
+    geom_line(data = line.df, aes(x = pooled.time, y = pb.counts, color = path), linetype = "solid", linewidth = 1, alpha = 0.7) +
     geom_line(data = curve.df, aes(x = x, y = y, color = path), linetype = "dashed", linewidth = 1, alpha = 0.7) +
     ggtitle(
       paste("Feature Id:", feature_id),

R/plotTrendCluster.R

@@ -48,6 +48,8 @@ plotTrendCluster <- function(scmpObj,
   # Global vars
   scmp_clusters <- "scmp_clusters"
   feature_id <- "feature_id"
+  # Offset
+  offset_vector <- scmpObj@Design@offset
 
   # Check
   assert_that(!isEmpty(scmpObj@Significant@clusters),
@@ -103,7 +105,7 @@ plotTrendCluster <- function(scmpObj,
           logs = log, logType = log_type,
           pseudoCount = pCount,
           significant = sig,
-          summary_mode = summary
+          summary_mode = summary,
         )
         return(plt)
       },
@@ -264,6 +266,15 @@ plotTrendCluster <- function(scmpObj,
     " (", curves_combined[["num"]], " Features)"
   )
 
+  if (sum(offset_vector) != 0) {
+    lines_combined <- points_combined
+  }
+
+
+  # View(points_combined)
+  # View(curves_combined)
+  # View(lines_combined)
+
   # Initiate plotting
   p <- ggplot() +
     geom_point(

R/sc.squeeze.R

@@ -76,6 +76,9 @@ sc.squeeze <- function(scmpObj,
                        aggregate = "sum",
                        fill_gaps = FALSE,
                        additional_params = list(use_unique_time_points = FALSE)) {
+  # suppressPackageStartupMessages(library(entropy))
+  # suppressPackageStartupMessages(library(assertthat))
+  # suppressPackageStartupMessages(library(SingleCellExperiment))
   # Initiate Variable
   scmp_bin_lower_bound <- "scmp_l_bound"
   scmp_bin_upper_bound <- "scmp_u_bound"
@@ -286,20 +289,12 @@ sc.squeeze <- function(scmpObj,
         # Increment the iteration counter
         it <- it + 1
       }
-    }
-
-    if (verbose) {
-      message(paste(
-        "Optimizing bin sizes, with maximum allowed bin size as",
-        max.allowed
-      ))
-    }
-
-    if (verbose) {
-      message(paste(
-        "Finally, for", path, ",", length_n, "time points has been compressed to",
-        nrow(bin_table), "bins and the sum is ", sum(bin_table[[bin.size]])
-      ))
+      if (verbose) {
+        message(paste(
+          "Finally, for", path, ",", length_n, "time points has been compressed to",
+          nrow(bin_table), "bins and the sum is ", sum(bin_table[[bin.size]])
+        ))
+      }
     }
 
     rownames(bin_table) <- NULL

vignettes/Basic-Workflow.Rmd

@@ -327,7 +327,7 @@ different expression between the two paths. Additionally, there are 10 genes
 uniquely associated with Path2vsPath1. This implies that Path2 has 10 genes that
 are significantly differentially expressed over time, using Path1 genes as a
 reference. Let's explore a few of these genes:
-```{r, "trend",eval=TRUE, echo=TRUE}
+```{r, "trend",eval=TRUE, echo=TRUE, fig.width=7, fig.height=7}
 FigureA <- plotTrend(scmp_ob, "Gene9", logs = TRUE, logType = "log")
 FigureB <- plotTrend(scmp_ob, "Gene95", logs = TRUE, logType = "log")
 FigureC <- plotTrend(scmp_ob, "Gene10", logs = TRUE, logType = "log")
@@ -339,7 +339,7 @@ These plots illustrate the gene expression trends for selected genes, providing
 insights into their behavior in different paths. Fortunately, since we are using
 simulated data, we can validate these trends by extracting gene-level metadata
 from our simulated object:
-```{r, "ground truth",eval=TRUE, echo=TRUE}
+```{r, "ground truth",eval=TRUE, echo=TRUE, fig.width=7, fig.height=7}
 groundTruth <- as.data.frame(splat.sim@rowRanges@elementMetadata)
 print(groundTruth[groundTruth$Gene %in% c("Gene9", "Gene95", "Gene10", "Gene92"), c(1, 2, 6, 8)])
 ```
@@ -382,6 +382,18 @@ further biological interpretation.
 This concludes the basic usage quick start guide of `scMaSigPro`. Please refer 
 to other vignettes for more in-depth analysis.
 
+```{r, "Save and Reload", eval=TRUE, echo=FALSE, message=FALSE, warning=FALSE}
+# Save the object
+ save(splat.sim, file = "../data/splat.sim.RData")
+
+# Save processed data for next tutorial
+scmp.ob <- scmp_ob
+save(scmp.ob, file = "../data/scmp.ob.RData")
+
+# Compress
+tools::resaveRdaFiles(paths = "../data/")
+```
+
 ---
 
 ### Session Info

vignettes/scMaSigPro-maSigPro.Rmd

@@ -394,7 +394,7 @@ plotTrendCluster(
   scmpObj = scmp_ob,
   plot = "coeff",
   verbose = FALSE
-)
+  )
 ```
 
 ---