Removed un-used commented codes and update vignette.

R/gs_score-methods.R

@@ -64,7 +64,7 @@ setMethod(
     ## compute score
     score <- gs_score(data = expr, features = features, suffix = suffix)
 
-    data@colData <- cbind(data@colData, score)
+    colData(data) <- cbind(colData(data), score)
 
     return(data)
   }

R/scale_mgm.R

@@ -13,7 +13,7 @@ scale_mgm <- function(expr, label) {
   sds <- sparseMatrixStats::rowSds(expr, na.rm = TRUE)
   # sds <- sapply(unique(label), \(i)
   #               sparseMatrixStats::rowSds(expr[, label == i], na.rm = TRUE)
-  #        ) # get mean of each group
+  #        ) # get sds of each group
   # colnames(sds) <- unique(label)
 
   ## compute group means

R/tf_idf_iae_wrappers.R

@@ -435,20 +435,6 @@ iae_hdb <- function(expr, features = NULL, multi = TRUE,
   ## factor cluster
   cluster <- factor(cluster)
 
-  # # thres <- 0
-  # # thres <- sparseMatrixStats::rowQuantiles(expr[features, , drop = FALSE], probs = 0.25, na.rm = TRUE)
-  # expr_offset <- expr[features, , drop = FALSE] - thres ## subtract offset
-  # expr_offset[expr_offset < 0] <- 0
-  #
-  # mean_row_in <- sapply(levels(cluster), function(type) {
-  #   rowMeans(expr_offset[, cluster == type, drop = FALSE], na.rm = TRUE)
-  # }) |> setNames(levels(cluster))  ## mean counts for each gene in the group
-  # mean_row_notin <- sapply(levels(cluster), function(type) {
-  #   apply(mean_row_in, 1, function(x) max(x[names(x) != type]))
-  # }) |> setNames(levels(cluster))  ## mean counts for each gene not in group
-  #
-  # iae <- log1p((mean_row_in/(mean_row_notin+0.01))[, cluster, drop = FALSE])  ## IDF scores
-
   iae <- iae_prob(
     expr = expr, features = features,
     label = cluster, multi = multi,

vignettes/smartid_Demo.Rmd

@@ -85,8 +85,8 @@ defac <- as.data.frame(rowData(data_sim)[, cols])
 up <- lapply(cols, \(id)
 dplyr::filter(defac, if_all(-!!sym(id), \(x) !!sym(id) / x > fc)) |>
   rownames())
-data_sim@metadata$up_markers <- setNames(up, cols)
-data_sim@metadata$up_markers
+slot(data_sim, "metadata")$up_markers <- setNames(up, cols)
+slot(data_sim, "metadata")$up_markers
 data_sim
 ```
 
@@ -112,12 +112,13 @@ idf_iae_methods()
 ```
 
 The basic version of TF, IDF and IAE can be termed as:
-$\mathbf{TF_{i,j}}=\frac{N_{i,j}}{\sum_j{N_{i,j}}}$
-$\mathbf{IDF_i} = log(1+\frac{n}{n_i+1})$
+
+$\mathbf{TF_{i,j}}=\frac{N_{i,j}}{\sum_j{N_{i,j}}},$
+$\mathbf{IDF_i} = log(1+\frac{n}{n_i+1}),$
 $\mathbf{IAE_i} = log(1+\frac{n}{\hat N_{i,j}+1})$
 
-$where\ N_{i,j}\ is\ the\ counts\ of\ feature\ i\ in\ cell\ j;\ \hat N_{i,j}\ is\ max(0,\ N_{i,j} - threshold);$
-$\ n\ is\ total\ counts\ of\ documents(cells);\ n_i\ is\ \sum_{j = 1}^{n} sign(N_{i,j} > threshold)$
+$\mathbf{where\ N_{i,j}\ is\ the\ counts\ of\ feature\ i\ in\ cell\ j;\ \hat N_{i,j}\ is\ max(0,\ N_{i,j} - threshold)};$
+$\mathbf{\ n\ is\ total\ counts\ of\ documents(cells);\ n_i\ is\ \sum_{j = 1}^{n} sign(N_{i,j} > threshold)}$
 
 Here for labeled data, we can choose logTF * IDF_prob * IAE_prob for marker identification.
 
@@ -174,7 +175,7 @@ It's clear that the real UP DEGs are popping up to the top n features. And for t
 score_barplot(
   top_markers = top_m,
   column = ".dot",
-  f_list = data_sim@metadata$up_markers,
+  f_list = slot(data_sim, "metadata")$up_markers,
   n = 20
 )
 ```
@@ -185,14 +186,14 @@ This can also be confirmed in data simulation information, where the scale facto
 
 ```{r}
 ggplot(data.frame(
-  "Gene76" = data_sim@metadata$tf["Gene76", ],
+  "Gene76" = slot(data_sim, "metadata")$tf["Gene76", ],
   Group = data_sim$Group
 )) +
   geom_violin(aes(x = Group, y = Gene76, fill = Group)) +
   theme_bw()
 
 ## sim gene info
-data_sim@rowRanges@elementMetadata[76, ]
+SummarizedExperiment::elementMetadata(data_sim)[76, ]
 ```
 
 ## Marker Selection
@@ -222,7 +223,7 @@ We can also compare our selected markers with real DEGs. As there is no markers
 ```{r}
 library(UpSetR)
 
-upset(fromList(c(data_sim@metadata$up_markers, marker_ls)), nsets = 6)
+upset(fromList(c(slot(data_sim, "metadata")$up_markers, marker_ls)), nsets = 6)
 ```
 
 `smartid` also provides some other implementation of marker selection. Here is another example using `mclust`. Different from `markers_mixmdl()`, `markers_mclust()` doesn't need a pre-defined number of components (which is 3 in `markers_mixmdl()`), instead, it will select the number of components by searching a series of potential numbers. This method is sometimes more robust than `markers_mixmdl()`.
@@ -258,7 +259,7 @@ Here we choose logTF * IDF_sd * IAE_sd for for gene-set scoring as a use case.
 
 $\mathbf{score}=logTF*IDF_{sd}*IAE_{sd}$
 
-$where\ \mathbf{IDF} = log(1+sd(N_{i})*\frac{n}{n_i+1})$
+$where\ \mathbf{IDF} = log(1+sd(N_{i})*\frac{n}{n_i+1}),$
 $\mathbf{IAE} = log(1+sd(N_{i})*\frac{n}{\sum_{j=1}^{n}N_{i,j}+1})$
 
 ## Score Samples
@@ -296,15 +297,15 @@ data_sim <- gs_score(
 )
 
 ## saved score
-colnames(data_sim@colData)
+colnames(colData(data_sim))
 ```
 
 Now we get 3 columns of score for each group markers. We can then visualize the score across groups, see how well it can discern the target group.
 
 It's evident that the score can sufficiently separate the target group from all others.
 
 ```{r, fig.width=10, fig.height=3}
-as.data.frame(data_sim@colData) |>
+as.data.frame(colData(data_sim)) |>
   tidyr::pivot_longer("Group1.score.unlabel":"Group3.score.unlabel",
     names_to = "group markers",
     values_to = "score"