hiReadsProcessor (version 1.8.2)

getIntegrationSites: Obtain integration sites from BLAT output

Description

Given a GRanges object from read.psl, the function uses specified filtering parameters to obtain integration sites and maintain sequence attrition. The function will remove any non-best scoring alignments from the object if not already filtered apriori.

Usage

getIntegrationSites(psl.rd = NULL, startWithin = 3,
  alignRatioThreshold = 0.7, genomicPercentIdentity = 0.98,
  correctByqStart = TRUE, oneBased = FALSE)

Arguments

psl.rd

a GRanges object reflecting psl format where tName is the seqnames.

startWithin

upper bound limit of where the alignment should start within the query. Default is 3.

alignRatioThreshold

cuttoff for (alignment span/read length). Default is 0.7.

genomicPercentIdentity

cuttoff for (1-(misMatches/matches)). Default is 0.98.

correctByqStart

use qStart to correct genomic position. This would account for sequencing/trimming errors. Position=ifelse(strand=="+",tStart-qStart,tEnd+qStart). Default is TRUE.

oneBased

the coordinates in psl files are "zero based half open". The first base in a sequence is numbered zero rather than one. Enabling this would add +1 to the start and leave the end as is. Default is FALSE.

Value

• a GRanges object with integration sites which passed all filtering criteria. Each filtering parameter creates a new column to flag if a sequence/read passed that filter which follows the scheme: 'pass.FilterName'. Integration Site is marked by new column named 'Position'.

See Also

startgfServer, read.psl, blatSeqs, blatListedSet, findIntegrations, pslToRangedObject, clusterSites, isuSites, crossOverCheck, read.blast8

Examples

data(psl)
psl.rd <- pslToRangedObject(psl)
getIntegrationSites(psl.rd)