mqppep_preproc: mqppep_anova_script.Rmd comparison

comparison mqppep_anova_script.Rmd @ 31:e103de3e41e6 draft

planemo upload for repository https://github.com/galaxyproteomics/tools-galaxyp/tree/master/tools/mqppep commit 7e8b616a2519c8faeb6fd743a8fb73f271f655f9

author	eschen42
date	Fri, 09 Dec 2022 21:06:08 +0000
parents	87794cf65bc0
children	072fe8228dfa

comparison

equal deleted inserted replaced

-:87794cf65bc0
+:e103de3e41e6
 - "Nick Graham^[ORCiD 0000-0002-6811-1941, University of Southern California: Los Angeles, CA, US]"
 - "Larry Cheng^[ORCiD 0000-0002-6922-6433, Rutgers School of Graduate Studies: New Brunswick, NJ, US]"
 - "Art Eschenlauer^[ORCiD 0000-0002-2882-0508, University of Minnesota: Minneapolis, Minnesota, US]"
 date:
 - "May 28, 2018"
-- "; revised June 23, 2022"
+- "; revised December 7, 2022"
 lot: true
 output:
 pdf_document:
 toc: true
 toc_depth: 2
 # look-up tables for kinase descriptions
 kinaseNameUprtLutBz2: "./kinase_name_uniprot_lut.tabular.bz2"
 kinaseUprtDescLutBz2: "./kinase_uniprot_description_lut.tabular.bz2"
 # should debugging trace messages be printed?
 showEnrichedSubstrates: FALSE
 # should debugging nb/nbe messages be printed?
 printNBMsgs:          FALSE
+# showld row-scaling be applied to heatmaps: "none" or "row"
+defaultHeatMapRowScaling: "none"
 # should debugging trace messages be printed?
 printTraceMsgs:       FALSE
 # when debugging files are needed, set debugFileBasePath to the path
-#   to the directory where they should be writtn
+#   to the directory where they should be written
 debugFileBasePath:    !r if (TRUE) NULL else "test-data"
 ---
 ```{r setup, include = FALSE, results = 'asis'}
 )
 )
 knitr::knit_exit()
 }
 )
+g_default_heatmap_row_scaling <-
+params$defaultHeatMapRowScaling
+if (
+!is.character(g_default_heatmap_row_scaling) ||
+!(g_default_heatmap_row_scaling %in% c("row", "none"))
+) {
+cat("invalid defaultHeatMapRowScaling (must be 'row' or 'none')")
+knitr::knit_exit()
+}
 # intensityHeatmapRows: 50
 # TODO Validate >> 0 < 75
 g_intensity_hm_rows     <- params$intensityHeatmapRows
 if (!is.integer(g_intensity_hm_rows) || g_intensity_hm_rows < 1) {
 max_nchar_rowname <- max(nchar(rownames(x)))
 max_nchar_colname <- max(nchar(colnames(x)))
 my_limit <- g_intensity_hm_rows
 my_row_cex_scale <- master_cex * 150 / nrow_x
+#ACE row cex shrink hack begin
+my_row_cex_scale <- master_cex * 150 / max(nrow_x, ncol_x)
+#ACE row cex shrink hack end
 my_col_cex_scale <- 3.0
+#ACE col cex shrink hack begin
+if (ncol_x > 40)
+my_col_cex_scale <- 3.0 * 40 / ncol_x
+#ACE col cex shrink hack end
 my_asterisk_scale <- 0.4 * my_row_cex_scale
 my_row_warp <- 1
 my_note_warp <- 2
 my_row_warp <- 1
 my_row_cex_asterisk <-
 g_intensity_hm_rows,    #   values of 50 and 75 worked well
 master_cex = 1.0,               # basis for text sizes
 margins = NULL,                 # optional margins (bottom, right)
 cellnote = NULL,                # optional matrix of character; dim = dim(m)
 adj = 0.5,                      # adjust text: 0 left, 0.5 middle, 1 right
+row_scaling =                   # should row-scaling be applied if possible?
+g_default_heatmap_row_scaling,
 ...                             # passthru to hm2plus or heatmap.2
 ) {
 use_heatmap_1 <- FALSE
 peptide_count <- 0
 # emit the heading for the heatmap
 }
 par(op)
 }
 # invoke hm_call inner function
-if (sum(rowSums(!is.na(m_hm)) < 2))
+if (row_scaling != "row" || sum(rowSums(!is.na(m_hm)) < 2))
 hm_call(
 m_hm,
 "none",
-"log(intensities), unscaled, unimputed, and unnormalized"
+"log(intensities), unimputed, and unnormalized"
 )
 else
 hm_call(
 m_hm,
 "row",
 {
 if (nrow(m_hm) > 1)
 hm_call(
 m_hm,
 "none",
-paste(
+"log(intensities), zero-imputed, unnormalized"
-"log(intensities), unscaled,",
-"zero-imputed, unnormalized"
-)
 )
 else
 cat("\nThere are too few peptides to produce a heatmap.\n")
 },
 error = function(r) {
 imp_smry_pot_peptides_after <- sum(good_rows)
 imp_smry_rejected_after  <- sum(!good_rows)
 imp_smry_missing_values_after   <- sum(is.na(quant_data_imp[good_rows, ]))
-# From ?`%in%`, %in% is currently defined as function(x, table) match(x, table, nomatch = 0) > 0
+# From ?`%in%`:
+#   %in% is currently defined as function(x, table) match(x, table, nomatch = 0) > 0
-sink(stderr())
-print("`%in%`:")
-print(`%in%`)
-sink()
 stock_in <-
 names(good_rows) %in%
 names(min_group_obs_count[g_intensity_min_per_class <= min_group_obs_count])
 if (print_nb_messages) nbe(see_variable(stock_in), "\n")
 grouping_factor =
 as.factor(grouping_factor$level), # anova_func arg2
 one_way_f = one_way_two_categories, # anova_func arg3
 simplify = TRUE # TRUE is the default for simplify
 )
-contrast_data_adj_p_values <- p.adjust(
-p = p_value_data_contrast_ps,
+if (!is.null(p_value_data_contrast_ps)) {
-method = "fdr",
+contrast_data_adj_p_values <-
-n = length(p_value_data_contrast_ps) # this is the default, length(p)
+p.adjust(
-)
+p = p_value_data_contrast_ps,
-# - compute the fold-change
+method = "fdr",
-contrast_p_df <-
+n = length(p_value_data_contrast_ps) # this is the default, length(p)
-data.frame(
+)
-contrast = contrast,
-phosphopep = contrast_cast$phosphopep,
+# - compute the fold-change
-p_value_raw = p_value_data_contrast_ps,
+contrast_p_df <-
-p_value_adj = contrast_data_adj_p_values
+data.frame(
-)
+contrast = contrast,
-db_write_table_overwrite <- (contrast < 2)
+phosphopep = contrast_cast$phosphopep,
-db_write_table_append <- !db_write_table_overwrite
+p_value_raw = p_value_data_contrast_ps,
-RSQLite::dbWriteTable(
+p_value_adj = contrast_data_adj_p_values
-conn = db,
+)
-name = "contrast_ppep_p_val",
+db_write_table_overwrite <- (contrast < 2)
-value = contrast_p_df,
+db_write_table_append <- !db_write_table_overwrite
-append = db_write_table_append
+RSQLite::dbWriteTable(
-)
+conn = db,
-# Create UK for insert
+name = "contrast_ppep_p_val",
-ddl_exec(db, "
+value = contrast_p_df,
-CREATE UNIQUE INDEX IF NOT EXISTS contrast_ppep_p_val__uk__idx
+append = db_write_table_append
-ON contrast_ppep_p_val(phosphopep, contrast);
+)
-"
+# Create UK for insert
-)
+ddl_exec(db, "
+CREATE UNIQUE INDEX IF NOT EXISTS contrast_ppep_p_val__uk__idx
+ON contrast_ppep_p_val(phosphopep, contrast);
+"
+)
+}
 }
 # Perhaps this could be done more elegantly using unique keys
 #   or creating the tables before saving data to them, but this
 #   is fast and, if the database exists on disk rather than in
 #   memory, it doesn't stress memory.
 hm_main_title
 = "Unnormalized (zero-imputed) intensities of enriched kinase-substrates",
 suppress_row_dendrogram = FALSE,
 master_cex              = 0.35,
 sepcolor                = "black",
-colsep                  = sample_colsep
+colsep                  = sample_colsep,
+row_scaling             = "none"
 )
 if (number_of_peptides_found > 1) {
 tryCatch(
 {

Mercurial > repos > eschen42 > mqppep_preproc

comparison mqppep_anova_script.Rmd @ 31:e103de3e41e6 draft