A tool to quantitatively assess differential translations with ribosome profiling data.

By pairwise comparisons of ribosome profiling data, Xtail identifies differentially translated genes across two experimental or physiological conditions.

xtail(
  mrna,
  rpf,
  condition,
  baseLevel = NA,
  minMeanCount = 1,
  normalize = TRUE,
  p.adjust.method = "BH",
  threads = NA,
  bins = 10000L,
  ci = 0
)

Arguments

mrna	a matrix or data frame of raw mRNA count data whose rows correspond to genes and columns correspond to samples. The column names should be non-empty, and in same order with condition.
rpf	a matrix or data frame of raw RPF count data whose rows correspond to genes and columns correspond to samples.The column names should be non-empty, and in same order with condition.
condition	condition labels corresponding to the order of samples in mrna and rpf. There must be exactly two unique values.
baseLevel	The baseLevel indicates which one of the two conditions will be compared against by the other one. If not specified, Xtail will return results of comparing the second condition over the first one.
minMeanCount	Xtail uses the average expression level of each gene, across all samples as filter criterion and it omits all genes with mean counts below minMeanCount.
normalize	Whether normalization should be done (TRUE \ FALSE). If missing, Xtail will perform median-of-ratios normlazation by default.
p.adjust.method	The method to use for adjusting multiple comparisons, by default "BH", see ?p.adjust
threads	The number of CPU cores used. By default, all available cores are used.
bins	The number of bins used for calculating the probability density of log2FC or log2R (default is 10000). This paramater will determine accuracy of pvalue. Set it small for a very quick test run.
ci	The level of confindence to get credible intervals of log2 fold change of translational efficiency (TE), for example 0.95.

Value

a xtail object

Details

No missing values are allowed in input data mrna and rpf.

Duplicate row names (gene names or gene ids) are not allowed.

Xtail takes in raw read counts of RPF and mRNA, and performs median-of-ratios normalization. Alternatively, users can provide normalized read counts and skip the built-in normal by setting "normalize" to FALSE.

The step of estimation of the probability distributions, for log2FC or log2R, will execute slowly in the current implementation, but can be speeded up by running on multiple cores using the parallel library. By default, the "detectCores" function in parallel library is used to determine the number of CPU cores in the machine on which R is running. To adjust the number of cores used, use "threads" argument to assign.

References

Zhengtao Xiao, Qin Zou, Yu Liu, and Xuerui Yang: Genome-wide assessment of differential translations with ribosome profiling data.

Author

Zhengtao xiao

Examples

#load the data
data(xtaildata)
# Get the mrna count data and rpf count data. For the example only the first
# 100 are used
test.mrna <- xtaildata$mrna[1:100,]
test.rpf <- xtaildata$rpf[1:100,]

#Assign condition labels to samples.
condition <- c("control","control","treat","treat")

#run xtail
test.results <- xtail(test.mrna,test.rpf,condition, threads = 2)
#> Calculating the library size factors
#> 1. Estimate the log2 fold change in mrna
#> 2. Estimate the log2 fold change in rpf
#> 3. Estimate the difference between two log2 fold changes
#> 4. Estimate the log2 ratio in first condition
#> 5. Estimate the log2 ratio in second condition
#> 6. Estimate the difference between two log2 ratios
#> Number of the log2FC and log2R used in determining the final p-value
#>  log2FC: 16
#>  log2R: 84
test.results
#> A xtail object:
#> Number of genes tested: 100 
#> Number of the log2FC and log2R used in determining the final p-value:
#>      log2FC: 16 
#>      log2R : 84 
#> 
#> Number of result with adjusted pvalue <  0.1 
#>      log2FC_TE > 0 (up)  : 0 
#>      log2FC_TE < 0 (down): 2