survFASE: Survival analysis using metafeatures

In harshsharma-cb/FASE: Analysis of RNA-Sequencing data using FASE (Finding Alternative Splicing Events).

Survival analysis using metafeatures

Description

survFASE finds survival rate of patients using altenative splicing data. It requires exon, intron and junction expression of the samples generated using FASE (check EPrnaseq, iPrnaseq and DEJ). rownames of all expression files and clinical data should have the same identifier/sample ID, otherwise survFASE would not be able to perform survival analysis. survFASE uses RMM/iMM to find metafeature(s) associated with the given exonID/intronID and incorporate the expression of those metafeatures with respective exonID/intronID.

Usage

survFASE(
  Time = Time,
  Status = Status,
  Gcount,
  clinical.data,
  rmm = NULL,
  imm = NULL,
  eventID,
  threshold = 0.6,
  designM,
  Groups
)

Arguments

Time	column name of survival time in clinical data. (default = Time)
Status	column name of survival status (alive/dead status) in clinical data. 1 = dead and 0 = alive. (default = Status)
Gcount	Gene count matrix for the given gene. It should contain raw meta-feature counts of only patients. Rownames of gcount should be unique sample IDs that can be mapped to the clinical data.
clinical.data	clinical data of patients. It should contain survival time (days to last follow up) and survival status (0/1) of the patients. Rownames of clinical data should be unique, non-repeating sample IDs that can be mapped to the Gcount.
rmm	readMembershipMatrix matrix of the gene. It contains association between exons and other metafeatures in a gene and is generated by default as RMM.Rdata using readMembershipMatrix function. RMM is required only for finding survival rate associated with a cassette exon event.
imm	intronMembershipMatrix matrix of the gene. It contains association between introns and other metafeatures in a gene. It is generated by default as iMM.Rdata using intronMembershipMatrix function. iMM is required only for finding survival rate associated with an intron retention event.
eventID	exonID or intronID of the AS event for survival analysis.
designM	design matrix
Groups	list of sample groups

Value

survFASE returns an overall p-value, concordance index and Cox-PH statistics. The overall p-value suggests whether or not the given exon/intron significantly affects patient survival. C-index signifies goodness-of-fit of the model. Cox-PH results show which of the metafeatures associated with the exon/intron affect survival rate and their statistical inferences like hazard-ratio, beta-coefficient, etc.

harshsharma-cb/FASE documentation built on Aug. 6, 2023, 1:37 a.m.