internal: jabba.gwalk

In mskilab/skitools: Various mskilab R utilties

jabba.gwalk

Description

Computes greedy collection (i.e. assembly) of genome-wide walks (graphs and cycles) by finding shortest paths in JaBbA graph.

takes in a jabba object and threshold for clusters and "quasi-reciprocal" junctions

Dumps JaBbA graph into json

Classify full set of dRanger rearrangements into "tiers" of confidence

(1) Tier 1 BPresult>0 and somatic_score>min.score1 (2) Tier 2 BPresult=0 and somatic_score>min.score1 (3) Tier 3 min.score2<=somatic_score<=min.score2 & tumreads>min.reads

Given GRanges of bins will aggregate around new bin width.

Usage

jabba.gwalk(jab, verbose = FALSE, return.grl = TRUE)

jgraph(jab, thresh_cl = 1e+06, all = FALSE, thresh_r = 1000, clusters = FALSE)

jab2json(jab, file, maxcn = 100, maxweight = 100)

karyotrack(kag, paths = NULL, col = "red", pad = 0)

ra_tier(
  dra,
  min.score1 = 10,
  min.score2 = 4,
  min.treads1 = 10,
  min.treads2 = 3,
  max.nreads = Inf
)

rebin(cov, binwidth, field = names(values(cov))[1], FUN = mean, na.rm = TRUE)

Arguments

jab	input jab object
file	output json file
kag	output of karyograph
paths	GRanges or GRangesList
cov	GRanges of binned genome-wide coverage
binwidth	new binwidth

Details

Takes karyograph and outputs gTrack +/- highlighting of one or more paths defined as GRanges or GRangesList (for multiple paths) Edges will only be highlighted when the exact interval pair corresponding to the edge is included in the graph

Value

GRangesList of walks with copy number as field $cn, cyclic walks denoted as field $is.cycle == TRUE, and $wid (width) and $len (segment length) of walks as additional metadata#'

gTrack of karyograph with particular nodes / edges colored with specified colors

GRanges of binned genome-wide coverage at new bin

Author(s)

Marcin Imielinski

Xiaotong Yao first peel off "simple" paths i.e. zero degree ends with >0 copy number so now we want to subtract that cn units of that path from the graph so we want to update the current adjacency matrix to remove that path while keeping track of of the paths on the stack then find paths that begin at a node and end at (one of its) immediate upstream neighbors this will be a path for whom col index is = parent(row) for one of the rows so now we want to subtract that cn units of that path from the graph so we want to update the current adjacency matrix to remove that path while keeping track of of the cycles on the stack

Marcin Imielinski ++ = RL +- = RR -+ = LL

mskilab/skitools documentation built on March 5, 2025, 11:52 p.m.