MACPET: An R-package for binding site analysis of ChIA-PET data.

In MACPET: Model based analysis for paired-end data

Description

The MACPET package can be used for general analysis of paired-end (PET) data like ChIA-PET. MACPET currently implements the following five stages: Linker filtering (stage 0), mapping to the reference genome (stage 1), PET classification (stage 2), peak-calling (stage 3) and interaction analysis (stage 4). The interaction analysis stage (4) currently supports only intra-chromosomal data. All of the MACPET stages can be run at once, or separately. In stage 0, MACPET identifies the linkers in the fastq files and classifies the reads as usable, chimeric or ambiguous. Usable reads are considered in the subsequent stages. In stage 1, MACPET maps the usable reads to the reference genome using bowtie and produces a paired-end BAM file. This BAM file is further used in stage 2 to classify the PETs as self-ligated/intra- or inter-chromosomal. Self-ligated PETs are used in stage 3 for the identification of significant peaks. In stage 3, MACPET segments the genome into regions and applies 2D mixture models for identifying candidate peaks using skewed generalized students-t distributions (SGT). It then uses a local poisson model for finding significant binding sites. Finally, in stage 4, MACPET uses the peaks from stage 3 as well as the Intra-chromosomal PETs from stage 2 to running an additive interaction-analysis model. MACPET is mainly written in C++, and it supports the BiocParallel package.

MACPET main function

MACPETUlt runs the whole analysis at once.

MACPET classes

PSelf

S4 class for Self-ligated PETs.

PSFit

S4 class for Self-ligated PETs after peak-calling.

PInter

S4 class for Inter-chromosomal PETs.

PIntra

S4 class for Intra-chromosomal PETs.

GenomeMap

S4 class for the interactions found in the data.

MACPET methods

plot

Method for plotting different objects.

summary

Method for summarizing different objects.

TagsToGInteractions

Method for converting Tags to GInteractions class.

PeaksToGRanges

Method for converting peaks to GRanges class.

exportPeaks

Method for exporting peaks in cvs file format.

ConvertToPSelf

Method for converting a GInteractions class of Self-ligated PETs to object of PSelf class.

GetSignInteractions

Method for subseting the significant interactions of a GenomeMap object.

GetShortestPath

Method for finding the shortest path between two peaks given a set of interactions.

PeaksToNarrowPeak

Method for converting peaks to narrowPeak (BED) format for use in interaction analysis using the MANGO/MICC algorithm. Alternatively the user can run stage 4 at MACPETUlt.

MACPET supplementary functions

ConvertToPE_BAM

Function for converting two separate BAM files into one paired-end BAM file.

AnalysisStatistics

Prints summary of multiple objects.

MACPET sample data

SampleChIAPETData.bam

Sample ChIA-PET data.

SampleChIAPETDataRead_1.bam

First reads from the sample ChIA-PET data.

SampleChIAPETDataRead_2.bam

Second reads from the sample ChIA-PET data.

MACPET_pinterData.rda

Sample PInter data.

MACPET_pintraData.rda

Sample PIntra data.

MACPET_pselfData.rda

Sample PSelf data.

MACPET_psfitData.rda

Sample PSFit data.

MACPET_GenomeMapData.rda

Sample GenomeMap data.

Author(s)

Ioannis Vardaxis, [e-mail]

References

Vardaxis I, Drabl<c3><b8>s F, Rye M and Lindqvist BH (2018). MACPET: Model-based Analysis for ChIA-PET. To be published.

Vardaxis I, Drabl<c3><b8>s F, Rye M and Lindqvist BH (2018). MACPET: Complete pipeline for ChIA-PET. To be published.

MACPET documentation built on Nov. 8, 2020, 5:47 p.m.