travis

Tools in R for the Analysis and Visualization of Intervals and Scores

Overview

travis is a suite of tools to process, analyze, and visualize bed and bedGraph data in R. Most of travis's functions can be categorized as either operating on beds or on bedGraphs.

Functions that operate on bed files:

• bedBreadth calculates the total number of bases covered in a bed file
• bedCat concatenates multiple bed files into one
• bedCenters takes a bed file and outputs a bed file of the midpoints of the intervals
• bedDivideMeta divides bed intervals (and optionally surrounding region) into an equal number of windows per interval, regardless of size
• bedDivide divides equally-sized bed invervals into an equal number of windows per interval
• bedEnds takes a bed file and outputs a bed file of the ends of the intervals
• bedHist plots a histogram of interval sizes
• bedOverhangs calculates frequencies of pairwise distances between the ends of intervals in a single bed file
• bedOverlap calculates overlap among two sets of bed files
• bedParseLengths parses bed intervals into separate files based on interval sizes
• bedParseStrands prases bed intervals into separate files based on strand
• bedPoissonTest perform a poisson test of interval counts between two bed files with a given set of windows
• bedRecenter alter the coordinates of a bed file
• bedRemoveChrom remove intervals on a specific chromosome
• bedSample randomly sample intevals of a bed file
• bedSizes load into R sizes of bed intervals
• bedSort sort a bed file by chromosome/coordinate
• bedStructures break up a bed12 file into separate components
• bedWords calculate nucleotide word frequencies for intervals in a bed file

Functions that operate on bedGraph files:

• bgAcf calculates autocorrelations of bedGraph files
• bgCorMatrix generates a matrix of pairwise correlations among a set of bedGraph files
• bgHist plots a histogram of bedGraph scores
• bgIqrNorm normalizes bedGraphs to have a specified interquartile range
• bgIqr calculate interquartile ranges of bedGraph files
• bgKsTest performs a KS test between two bedGraph files
• bgLoess loess smooths a bedGraph files
• bgOps perform various operations on single bedGraph files (e.g. transformations) or multiple bedGraph files (averages, differences, variance, ratios)
• bgParseScores parses bedGraph intervals into multiple bedGraphs based on score
• bgPlot plot bedGraph scores of specified intervals
• bgPoissonTest perform a poisson test between two bedGraph files
• bgQuantileNorm normalizes bedGraphs with quantile normalization
• bgQuantiles calculates specified quantiles of bedGraph files
• bgRead reads bedGraph files into R
• bgScatterGrid calculate a grid of pairwise density scatter plots using a set of bedGraph files
• bgScores read bedGraph scores into R
• bgSort sort bedGraph files by score
• bgThreshold outputs bedGraphs containing intervals exceeding a specified score
• bgUnify make multiple bedGraph files share an identical set of intervals
• bgZtest perform a Z-test between two bedGraph files

There are several functions that use bam or sam files:

• bamCount counts the number of reads in bam files
• bamParseLengths parses paired-end reads into multiple files based on insert size
• samStats calcaultes and plots summary statistics for alignment data in sam files

Finally, there are three plotting functions that operate directly on R objects:

• intervalBox draws a boxplot of data that is binned based on its score in another dimension
• rageHist plots overlayed histograms of data organized in lists or data frames
• scatterdens generates a density plot of points in a scatter plot, useful for generating a scatter plot of many points

Installation

travis heavily depends on bedtools. Other programs travis uses include several GNU programs that are typically preinstalled in most Linux distributions, including awk, sed, sort, shuf, and cut. For older linux distributions, you may need to install a more recent version of GNU core utilities that supports parallelization in sort.

In addition, there are R dependencies. Install devtools if not installed already:

install.packages("devtools")

Then install conifur (convenience functions for R), converge (conversion tools for genomic data), gyro (wrappers for genomic tools), and travis:

devtools::install_github("dvera/conifur")
devtools::install_github("dvera/converge")
devtools::install_github("dvera/gyro")
devtools::install_github("dvera/travis")