faseq: DASeq
In fchen365/surf: Statistical Utility for RBP (RNA binding protein) Functions

Description Usage Arguments Value References

Perform the functional association analysis (DASeq) in a single command. This function is a wrapper that calls the necessary functions in order for DASeq.

faseq(
  event,
  sampleData,
  signal.type = "FPKM",
  FUN.aggregate = "mean",
  minMQS = 10,
  minOverlap = 12,
  isPairedEnd = TRUE,
  cores = max(1, detectCores() - 2),
  ...,
  min.size = 100,
  trim = 0.025,
  fdr.cutoff = 0.05,
  signal.cutoff = 20,
  verbose = FALSE
)

`event`	a `surf` object.
`sampleData`	`data.frame`, must contain two columns – "bam" and "condition" (for "IP" and "input", where "IP" should come first), whose `row.names` represent the sample names. "bam" is the file name of CLIP-seq bam. "condition" will be coerced to factor, whose first level will be treated as IP, and the second level as input.
`signal.type`	`character`, indicate the type of feature signal wanted, support "TPM" for Transcripts Per Kilobase Million, "FPKM" for Fragments Per Kilobase Million (for paired-end reads) and Reads Per Kilobase Million (for single-end reads), and "raw.count" for raw read counts
`FUN.aggregate`	`function`, used for aggregating signals within `condition`, default to `mean()`.
`minMQS`	parameters for featureCounts. `minMQS` is default to 10, and `minOverlap` is default to 12 (25% of the typical read length of CLIP-seq (~50bp)), and `isPairedEnd` is default to `TRUE`.
`minOverlap`	parameters for featureCounts. `minMQS` is default to 10, and `minOverlap` is default to 12 (25% of the typical read length of CLIP-seq (~50bp)), and `isPairedEnd` is default to `TRUE`.
`isPairedEnd`	parameters for featureCounts. `minMQS` is default to 10, and `minOverlap` is default to 12 (25% of the typical read length of CLIP-seq (~50bp)), and `isPairedEnd` is default to `TRUE`.
`cores`	`integer`, number of available workers, sent to `nthreads` of featureCounts
`...`	parameters for Rsubread::featureCounts.
`min.size`	`integer`, the minimum size of "reliable" training set, default to 60.
`trim`	`numeric`, the percentile used to trim the training data. This is useful in producing a more robust estimation of functional association.
`fdr.cutoff`	`numeric`, significance cutoff for the adjusted p-values.
`signal.cutoff`	`numeric`, threshold cut-off for the eCLIP signals, default to 20. Set this to 0 if don't want to filter those location with low eCLIP signals of the RBP.
`verbose`	`logical`, whether (default to `TRUE`) to echo progress

a surf object DASeq results updated.

Chen, F., & Keles, S. (2020). SURF: integrative analysis of a compendium of RNA-seq and CLIP-seq datasets highlights complex governing of alternative transcriptional regulation by RNA-binding proteins. Genome Biology, 21(1), 1-23.

fchen365/surf documentation built on June 18, 2021, 12:02 p.m.