JACUSA - DEPRECATED

JAVA framework for accurate SNV assessment

---

JACUSA1 has been deprecated and superseded with JACUSA2. Likewise JACUSAhelper has been deprecated and replaced with JACUSA2helper. Please change to the new versions of the software.

Find source code and tools in the following sub-directories of the repository:

• src/ The main Java source code for JACUSA
• manual/manual.pdf The manual for JACUSA
• JacusaHelper R package to process JACUSA output file(s)
• tools/AddVariants Java tool to implant variants into BAM file

Requirements

JACUSA has been developed and tested with Java v1.8.

IMPORTANT! Stranded paired-end data are handled properly with JACUSA v1.2.0 and higher. DO NOT USE JACUSA v1.0.1 on stranded paired-end data!

Download

Get the current Jacusa JAR:

https://github.com/dieterich-lab/JACUSA/releases/download/v1.3.5/JACUSA_v1.3.5.jar

Changes:

v1.3.0

• Fixed an issue with call-1 and stranded reads. JACUSA now requires BAM files to contain MD field.
• Changed calculation of up/downstream matches between filtered features
• Changed Homopolymer calculation -> simple homopolymer within alignment blocks (insertions and introns currently ignored)

---

v1.2.3

• Fixed pileup filter to maintain orientation for all library type combinations
• Fixed typo in library type: FR_FIRSTSTRAND -> RF_FIRSTSTRAND

v1.2.0

• Added support for stranded paired end reads - parameter -P changed
• Added support for single sample mode
• Added -R | --SHOW-REF option
• Minor fixes / typos

v1.0.1

• Minor fixes / typos.

Older releases:

  DO NOT USE JACUSA v1.0.1 on stranded paired-end data!
https://github.com/dieterich-lab/JACUSA/raw/master/build/JACUSA_v1.0.1.jar

Important Change

Since v1.2 the format of -P has changed! The format has been inspired by tophat's http://ccb.jhu.edu/software/tophat/manual.shtml library type parameter. With the command line parameter -P,--build-pileup the user can choose from combinations of:

• FR-FIRSTSTRAND STRANDED library - first strand sequenced,
• FR-SECONDSTRAND STRANDED library - second strand sequenced, and
• UNSTRANDED UNSTRANDED library.

Usage

Available methods for JACUSA $ java -jar jacusa.jar [ENTER]:

• call-1 Call variants - one sample
• call-2 Call variants - two samples
• pileup SAMtools like mpileup for two samples

Single sample mode: call-1

General command line structure for variant calling call-1:

jacusa.jar call-2 [OPTIONS] BAM1_1[,BAM1_2,BAM1_3,...]

Get available options:

java -jar jacusa.jar call-1

Two sample mode: call-2

General command line structure for variant calling call-2:

jacusa.jar call-2 [OPTIONS] BAM1_1[,BAM1_2,BAM1_3,...] BAM2_1[,BAM2_2,BAM2_3,...]

Get available options:

java -jar jacusa.jar call-2

Example gDNA vs. cDNA (two sample mode)

Download and extract sample data

# goto to https://data.dieterichlab.org/s/hg19_chr1_gDNA_VS_cDNA
# download hg19_chr1_gDNA_VS_cDNA.tar.gz
# and unpack with
tar xzvpf hg19_chr1_gDNA_VS_cDNA.tar.gz

Call RNA-DNA differences (RDDs) by comparing gDNA and cDNA in sample data and save results in rdds.out.

$ java -jar call-2 -P UNSTRANDED,FR-FIRSTSTRAND -a H,M,B,Y -f 1024 -T 2.3	-p 2 -r rdds.out gDNA.bam cDNA1.bam,cDNA2.bam

JacusaHelper

Read, Process, and write JACUSA output files

Installation

Download JacusaHelper:

$ wget https://github.com/dieterich-lab/JACUSA/raw/master/JacusaHelper/build/JacusaHelper_0.43.tar.gz

Install JacusaHelper in R:

install.packages("JacusaHelper_0.43.tar.gz")
library("JacusaHelper")

Example

Load JacusaHelper package in R:

library("JacusaHelper")

Read JACUSA output, filter sites where the variant base is NOT present in all replicates of at least one sample, and finally add editing frequency info:

# Read Jacusa output and filter by test-statistic >= 1.56 and 
# ensure that site have at least 10 reads in (cov1) sample 1 and at least 5 reads per replicate in (covs2) sample 2
data <- Read("Jacusa_RDD.out, stat = 1.56, fields = c("cov1", "covs2"), cov = c(10, 5))
# This ensures that the variant base is present in all replicates of at least one sample
data <- FilterResult(data)
# This is only applicable for RDD calls and it will calculate their editing frequency.
# It is expected that gDNA is stored as sample 1!
data <- AddEditingFreqInfo(data)

Plot base change conversion:

# Among other additional infos, AddEditingFreqInfo will populate baseChange field in data
tbl <- table(data$baseChange)
barplot(tbl)

Check documentation in R for more details

?JacusaHelper

AddVariants

Add variants to a BAM file

Download

Get the current AddVariants JAR:

$ wget https://github.com/dieterich-lab/JACUSA/raw/master/tools/AddVariants/build/AddVariants_v0.3.jar

Usage

Implant variants defined in <input.bam> into <variants.bed> and write results to <output.sam>:

java -jar AddVariants.jar <input.bam> <variants.bed> | samtools view -Sb - > <output.sam>

Format of variants.bed

chr | start | end

License

see LICENSE file