The input dataframe contains 4,478 cells × 62 SNPs.
Distribution plots
Plot a distribution for every column, with the number of cells on the y-axis.
For each column, count values at both ends of the distribution: values equal to 0 or 100, values ≤1 or ≥99, values ≤2 or ≥98, and so on.
Define a tail-ratio function for each SNP
For lower = 0 and upper = 100, count values ≤0 or ≥100 in each column and divide by the number of values to obtain a ratio.
column_ValueOutOfRangeRatio <- function(matrix,
lower_criteria = 0,
higher_criteria = 100
){
result = data.frame(row.names = colnames(matrix))
result[,1] = 0
colnames(result) = paste0("Ratio_",lower_criteria,"_",higher_criteria)
# for loop for each SNPs
for (i in 1:ncol(matrix)){
count = sum(matrix[,i] <= lower_criteria | matrix[,i] >= higher_criteria)
ratio = round((count/nrow(matrix))*100, digits = 2)
# Find the row_index of the current SNPs in the data frame
row_index <- match(colnames(matrix)[i], rownames(result))
result[row_index, colnames(result)] <- ratio
}
return(result)
}Evaluate several lower and upper cutoffs
AF_62_mtx_2N <- AF_62_mtx_2N[,-c(1:2)]
# for loop for different range (i+j has to be 100)
SNP_AlleleFreqOutOfRangeRatios <- data.frame(row.names = colnames(AF_62_mtx_2N))
for (i in 0:10) {
j = 100 - i
new_df <- column_ValueOutOfRangeRatio(AF_62_mtx_2N,
lower_criteria = i,
higher_criteria = j)
SNP_AlleleFreqOutOfRangeRatios = cbind(SNP_AlleleFreqOutOfRangeRatios, new_df)
}Compare tail ratios with a line chart
Code
SNP_AlleleFreqOutOfRangeRatios$SNPs <- rownames(SNP_AlleleFreqOutOfRangeRatios)
# conver to long format
SNP_AlleleFreqOutOfRangeRatios_long <- tidyr::pivot_longer(data = SNP_AlleleFreqOutOfRangeRatios,
cols = -c(SNPs),
names_to = "range",
values_to = "ratio")
# make sure the order of x axis (range character)
SNP_AlleleFreqOutOfRangeRatios_long$range <- factor(SNP_AlleleFreqOutOfRangeRatios_long$range,
levels = unique(SNP_AlleleFreqOutOfRangeRatios_long$range))
####################################################################################
# Visualization
pdf(file = "PDF/QC/2N_SNP_AlleleFreqOutOfRangeRatios.pdf", width = 10, height = 6)
# chr3, 6, 7, 8
ggplot(SNP_AlleleFreqOutOfRangeRatios_long[grep("chr3|chr6|chr7|chr8",
SNP_AlleleFreqOutOfRangeRatios_long$SNPs,
ignore.case = TRUE),],
aes(x = range,
y = ratio,
group = SNPs)) +
geom_line(aes(color = SNPs))+
geom_point(aes(color = SNPs))+
theme_classic() +
theme(text = element_text(size=15),
axis.text.x = element_text(angle = 45, vjust = 1, hjust=1))+
ggtitle("Ratios of allele frequency that are out of range: chr3,chr6,chr7,chr8")
# chr10
ggplot(SNP_AlleleFreqOutOfRangeRatios_long[grep("chr10",
SNP_AlleleFreqOutOfRangeRatios_long$SNPs,
ignore.case = TRUE),],
aes(x = range,
y = ratio,
group = SNPs)) +
geom_line(aes(color = SNPs))+
geom_point(aes(color = SNPs))+
theme_classic() +
theme(text = element_text(size=15),
axis.text.x = element_text(angle = 45, vjust = 1, hjust=1))+
ggtitle("Ratios of allele frequency that are out of range: chr10")
# chr11, 12, 13
ggplot(SNP_AlleleFreqOutOfRangeRatios_long[grep("chr11|chr12|chr13",
SNP_AlleleFreqOutOfRangeRatios_long$SNPs,
ignore.case = TRUE),],
aes(x = range,
y = ratio,
group = SNPs)) +
geom_line(aes(color = SNPs))+
geom_point(aes(color = SNPs))+
theme_classic() +
theme(text = element_text(size=15),
axis.text.x = element_text(angle = 45, vjust = 1, hjust=1))+
ggtitle("Ratios of allele frequency that are out of range: chr11,chr12,chr13")
# chr14
ggplot(SNP_AlleleFreqOutOfRangeRatios_long[grep("chr14", SNP_AlleleFreqOutOfRangeRatios_long$SNPs, ignore.case = TRUE),],
aes(x = range,
y = ratio,
group = SNPs)) +
geom_line(aes(color = SNPs))+
geom_point(aes(color = SNPs))+
theme_classic() +
theme(text = element_text(size=15),
axis.text.x = element_text(angle = 45, vjust = 1, hjust=1))+
ggtitle("Ratios of allele frequency that are out of range: chr14")
# chr15
ggplot(SNP_AlleleFreqOutOfRangeRatios_long[grep("chr15",
SNP_AlleleFreqOutOfRangeRatios_long$SNPs,
ignore.case = TRUE),],
aes(x = range,
y = ratio,
group = SNPs)) +
geom_line(aes(color = SNPs))+
geom_point(aes(color = SNPs))+
theme_classic() +
theme(text = element_text(size=15),
axis.text.x = element_text(angle = 45, vjust = 1, hjust=1))+
ggtitle("Ratios of allele frequency that are out of range: chr15")
# chr16, 17
ggplot(SNP_AlleleFreqOutOfRangeRatios_long[grep("chr16|chr17",
SNP_AlleleFreqOutOfRangeRatios_long$SNPs,
ignore.case = TRUE),],
aes(x = range,
y = ratio,
group = SNPs)) +
geom_line(aes(color = SNPs))+
geom_point(aes(color = SNPs))+
theme_classic() +
theme(text = element_text(size=15),
axis.text.x = element_text(angle = 45, vjust = 1, hjust=1))+
ggtitle("Ratios of allele frequency that are out of range: chr16,chr17")
# chr18, 19
ggplot(SNP_AlleleFreqOutOfRangeRatios_long[grep("chr18|chr19",
SNP_AlleleFreqOutOfRangeRatios_long$SNPs,
ignore.case = TRUE),],
aes(x = range,
y = ratio,
group = SNPs)) +
geom_line(aes(color = SNPs))+
geom_point(aes(color = SNPs))+
theme_classic() +
theme(text = element_text(size=15),
axis.text.x = element_text(angle = 45, vjust = 1, hjust=1))+
ggtitle("Ratios of allele frequency that are out of range: chr18,chr19")
# chr21
ggplot(SNP_AlleleFreqOutOfRangeRatios_long[grep("chr21",
SNP_AlleleFreqOutOfRangeRatios_long$SNPs,
ignore.case = TRUE),],
aes(x = range,
y = ratio,
group = SNPs)) +
geom_line(aes(color = SNPs))+
geom_point(aes(color = SNPs))+
theme_classic() +
theme(text = element_text(size=15),
axis.text.x = element_text(angle = 45, vjust = 1, hjust=1))+
ggtitle("Ratios of allele frequency that are out of range: chr21")
# chr22, 25
ggplot(SNP_AlleleFreqOutOfRangeRatios_long[grep("chr22|chr25",
SNP_AlleleFreqOutOfRangeRatios_long$SNPs,
ignore.case = TRUE),],
aes(x = range,
y = ratio,
group = SNPs)) +
geom_line(aes(color = SNPs))+
geom_point(aes(color = SNPs))+
theme_classic() +
theme(text = element_text(size=15),
axis.text.x = element_text(angle = 45, vjust = 1, hjust=1))+
ggtitle("Ratios of allele frequency that are out of range: chr22,chr25")
dev.off()Each line represents one column (one SNP).
The y-axis is the proportion outside the chosen range; the x-axis identifies the range. For example, Ratio_10_90 is the proportion of values ≤10 or ≥90.
2N_SNP_AlleleFreqOutOfRangeRatios.pdf
Summary
The curves compare the tail proportions across SNPs and show how rapidly those proportions increase as the central interval narrows. The SNP chr10:40320637:A/C−10q has a notably different pattern.
Its histogram below also shows an asymmetric distribution.
