brucemoran / gintools

[Dev] Tools for genomic DNA integration analysis

gintools

Tools used for analysis and interpretation of foriegn DNA integration elements, such as retroviruses, retroviral-based gene therapy vectors, retrotransposons, or DNA oligos incorporated into a host genome by a Non-Homologous End Joining pathway.

Functions contained in the gintools package can be separated into different catagories:

• Nucleotide sequence processing - managing nucleotide sequence data from machine output, including demultiplexing, trimming, filtering, and consolidating. Alignment can be handled by the user's preferred sequence aligner (BLAT, BWA, ...).

• Alignment interpretation - interpret alignment information from PSL or SAM/BAM inputs. Identify important locations, standarize across samples, or resolve edges of observed nucleotide alignments to help reduce noise inherent in the data from PCR amplification and sequencing.

• Analytics - functions designed to assist in analysis of processed data. Track observations across specimens, determine abundances of clones, or consolidate data into a summary of observations.

• Utilities - functions designed to assist with the other catagories and serve to make seemingly simple operations just that.

Install

To install gintools, simply run the following command within an R session:

devtools::install_github("https://github.com/cnobles/gintools.git")

# Or

devtools::install_github("cnobles/gintools")

Functions by catagory

Nucleotide sequence processing

• banmat : A binary ambiguous nucleotide matrix based on NUC4.4.

Alignment interpretation

• standardize_sites : Returns a GRanges object where the site positions have been standardized with all other sites in the dataset which are within the gap distance.
• refine_breakpoints : Returns a GRanges object where the break point positions have been adjused based on positional clusterting and read counts within the dataset and specified distances.

Analytics

• track_clones : Returns a GRangesList of integration sites shared between multiple GRanges objects.
• condense_intsites : Returns a GRanges object containing a single integration site in each row, removing all breakpoint information.
• determine_abundance : Returns a data.frame with position ids and abundances, calculated by the number of unique fragment lengths or utilizing the sonicLength package.
• cluster_kv : Returns the cluster or group membership based on connections between key nodes based on supplied values.

Utilities

• generate_posid : Creates a character vector of position IDs in the format of [chromosome(+/-/*)position] given a GRanges object or ID information.
• db_to_granges : Converts an INSPIIRED database query into a GRanges object.
• unique_granges : Considers and keeps metadata columns when identifying unique ranges within a GRanges object.
• pop_calcs : Calculations for describing features of populations, i.e. Shannon Diversity, Clonality, Gini Index ...
• vzip : Combines two or more vectors in a "zipping" fashion and returns a single vector.
• vintersect : Identify intersecting values in two or more vectors.
• vcollapse : Collapse row contents into a single vector from a data.frame or matrix.

Dependencies

The gintools package depends on R 3.2 or higher and will Import the following packages during installation unless they are already present:

• dplyr (>= 0.7)
• GenomicRanges (>= 1.26)
• igraph (>= 1.0.1)
• IRanges (>= 2.10)
• magrittr(>= 1.0)
• Matrix (>= 1.2)
• parallel(>= 3.2)
• S4Vectors (>= 0.12)

The following packages are suggested for increased utility:

• geneRxCluster
• sonicLength
• testthat