test_candidates: Determines whether a candidate SNP has allele specific...

In angelalica/ASTranslation: Assessing Whether a SNP Affects Translation through Determination of Allele-Specific Translation

Description

This function helps the user determine whether a particular heterozygous SNP has allele-specific translation based on data from an assay that combines polysome profiling and digital droplet PCR (ddPCR). It uses bootstrapping techniques to determine whether the amount of mutant and wildtype transcripts in each fraction (directly from the polysome profiling or grouped by number of ribosomes) is significantly different from the expected values. Each fraction corresponds to a particular weight (heavier transcripts contain a greater number of ribosomes, indicating greater translation).

Usage

test_candidates(file.name, gene.name, expected.value = NULL,
  grouped = FALSE, grouped.fractions = NULL, zoom.range = NULL,
  nsims = 1000, output.pvals = "output_pvals.csv",
  output.plots = "output_plots.pdf")

Arguments

file.name	Filename of the ddPCR counts data file to be read in. The data must be formatted in the following manner: each row corresponds to a ddPCR well, and there must be exactly 7 columns - well ID, fraction ID, Ch1+/Ch2+ counts, Ch1+/Ch2- counts, Ch1-/Ch2+ counts, Ch1-/Ch2- counts, and proportion of total concentration.
gene.name	A string indicating the name of gene being studied.
expected.value	A number indicating the overall Ch1+/Ch2+ ratio that would be expected if there were no allele specific translation. The user can include if this value is already known. If NULL, the function will automatically calculate an expected value by summing over all fractions. Defaults to NULL.
grouped	A boolean indicating whether to group fractions by number of polysomes. Defaults to FALSE.
grouped.fractions	A vector indicating the number of experimental fractions corresponding to each ribosomal fraction. Defaults to NULL.
zoom.range	A vector of 2 numbers that indicate the particular range of experimental fractions to graph. Both numbers must be positive. The first number must be strictly smaller than the second number and neither can be greater than the total number of fractions. Defaults to NULL, which corresponds to plotting all fractions. NOTE: zoom.range can only be used with fractions that are NOT grouped by ribosomes.
nsims	The number of bootstrap simulations to run for each fraction
output.pvals	File name of the csv output file
output.plots	File name of the pdf output file

Value

There are two outputs. The first is a CSV file with each fraction / group, it's p-value, and its proportion of the total concentration. The second output is a PDF containing two aligned graphs. One graph corresponds to a boxplot of all bootstrapped Ch1+/Ch2+ ratios relative to the expected value (either calculated or inputted by the users). This makes it visually clear which fractions are significantly different (p < 0.05 for a two-tailed test) from the expected ratios. The other graph is a barplot that graphs the ddPCR concentrations from each fraction so that users can see the distribution and relative concentrations of RNA across all fractions (from the early fractions with no ribosomes to the heavier fractions with 5+ ribosomes).

Examples

  test_candidates (filename = "gene1_ddPCR_data.csv", gene.name = "GENE1")
  test_candidates (filename = "gene2_ddPCR_data.csv", gene.name = "GENE2", zoom.range = c(5, 16))
  test_candidates (filename = "gene3_ddPCR_data.csv", gene.name = "GENE3", expected.value = 1.2, grouped = TRUE, grouped.fractions = c(7, 3, 1, 1, 1, 5))

angelalica/ASTranslation documentation built on May 10, 2019, 11:46 a.m.