20251125

2025-11-25 00:28:51 +08:00
commit eb3f16c30e
406 changed files with 91653 additions and 0 deletions
@@ -0,0 +1,118 @@
+
+plot_log2_fpkm_dist = function (fpkm_files) {
+
+    num_files = length(fpkm_files);
+
+	data_list = list();
+	
+	max_y = 0
+	xlim = c(0,0)
+
+	for (i in 1:num_files) {
+
+		file = fpkm_files[i]
+      	data = read.table(file, header=F)
+		data = data[,6];
+		data = log2(data+1)
+		
+		den = density(data);
+	
+		data_list[[i]] = den
+
+		y = max(den$y);
+		if (y > max_y) {
+			max_y = y;
+		}
+	
+		x = min(den$x)
+		if (x < xlim[1]) {
+			xlim[2] = x
+		}	
+
+		x = max(den$x);
+		if (x > xlim[2]) {
+			xlim[2] = x
+		}
+
+
+	}
+	
+	colors = rainbow(num_files);
+
+	for (i in 1:num_files) {
+
+		if (i == 1) {
+
+			plot(data_list[[1]], col=colors[1], xlim=xlim, ylim=c(0,max_y), xlab="log2(fpkm+1)")
+		}
+		else {
+			points(data_list[[i]], col=colors[i], type='l')
+		}
+	}
+
+	return;
+}
+
+
+
+
+
+plot_expressed_gene_counts = function(fpkm_file, 
+                                      title="expressed transcript counts vs. min fpkm", 
+                                      fpkm_range=seq(0,10,0.2), 
+                                      total=0, 
+                                      outfile="count_summary.txt") {
+
+    data = read.table(fpkm_file, header=T, row.names=1);
+	data = data[,5]
+	
+	
+	counts_expressed = c();
+    counts_not_expressed = c();
+
+    print_not_expressed_flag = 1;
+    if (total == 0) {
+        total = length(data);
+        print_not_expressed_flag = 0;
+    }
+
+    count_expressed = total;
+	for (i in fpkm_range) {
+   
+        if (i > 0) {
+       		count_expressed = sum(data>=i)
+		}
+        count_not_expressed = total - count_expressed;
+	
+		counts_expressed[length(counts_expressed)+1] = count_expressed;
+		counts_not_expressed[length(counts_not_expressed)+1] = count_not_expressed;
+		
+	}
+
+
+    orig_settings = par(mfrow=c(1,2))
+
+	plot(fpkm_range, counts_expressed, type='o', col='black', xlab="min(fpkm)", 
+        main=title, ylab="count of transcripts", ylim=c(0,max(counts_expressed, counts_not_expressed)))
+    
+    if (print_not_expressed_flag) {
+    	points(fpkm_range, counts_not_expressed, type='o', col='blue')
+        legend('topright', c('expressed', 'not expressed'), col=c('black', 'blue'), pch=15);
+    }   
+	else {
+		legend('bottomright', c('expressed transcripts'), col=c('black'), pch=15);
+	}
+   
+	data_table = data.frame(fpkm_range=fpkm_range, expressed=counts_expressed, not_expressed=counts_not_expressed);
+	
+	write.table(data_table, file=outfile, quote=F, sep='\t', row.names=F);
+	
+
+    ## make density plot for non-zero FPKM values
+    data = data[data>0]
+    plot(density(log2(data)), xlab="log2(fpkm)")
+
+
+    par(orig_settings)
+    
+}
@@ -0,0 +1,102 @@
+
+get_Poisson_conf_intervals = function(seq_range, quantile_vec = c(0.05,0.95), plot=T) {
+	
+	num_rows = length(seq_range)
+	num_cols = length(quantile_vec)
+	m = matrix(nrow=num_rows, ncol=num_cols)
+	
+    colnames(m) = quantile_vec	
+	rownames(m) = seq_range
+
+	row_count = 0
+	for (i in seq_range) {
+		q = qpois(quantile_vec, i)
+		row_count = row_count + 1
+		m[row_count,] = q			
+	
+	}
+
+    results = list()
+    results$mat = m
+	
+	if (plot) {
+		percents = plot_conf_intervals(m)
+ 	    
+        results$pct = percents
+    }
+
+	return(results)
+
+}
+
+
+get_NB_conf_intervals = function(seq_range, dispersion, quantile_vec = c(0.05,0.95), plot=T) {
+	
+    size = 1/dispersion  #according to the mu-definition of size in nbinom of R, where var = mean + (1/size)mean^2
+    
+	num_rows = length(seq_range)
+	num_cols = length(quantile_vec)
+	m = matrix(nrow=num_rows, ncol=num_cols)
+	
+    colnames(m) = quantile_vec	
+	rownames(m) = seq_range
+
+	row_count = 0
+	for (i in seq_range) {
+		q = qnbinom(quantile_vec, mu=i, size=size)
+		row_count = row_count + 1
+		m[row_count,] = q			
+	
+	}
+	
+    results = list()
+    results$mat = m
+    
+
+	if (plot) {
+		percents = plot_conf_intervals(m)
+ 	
+        results$pct = percents
+    }
+
+	return(results)
+
+}
+
+
+
+
+
+
+plot_conf_intervals = function(m) {
+
+	par(mfrow=c(1,2))
+
+	c_names = colnames(m)
+	r_vals = as.numeric(rownames(m))
+
+	max_val = max(m)
+	
+	plot(r_vals, r_vals, xlab="known read counts", ylab="Poisson read counts dist", t='l')
+
+	# plot the confidence levels
+	line_colors = rainbow(length(c_names))
+	for(i in 1:length(c_names)) {
+		points(r_vals,m[,i], t='l', col=line_colors[i])
+	}
+
+	# plot the max percentage of value for 95% conf level.
+	percents=c()
+	for (i in 1:length(r_vals)) {
+
+		max_delta = max(abs(m[i,]-r_vals[i]))
+		percent = max_delta/r_vals[i]*100
+		
+		percents[i]=percent;
+	}
+	
+	plot(r_vals, percents, ylim=c(0,100), xlab="read counts", ylab="percent of value for 95% conf interval")
+
+	percents
+
+}