RADIA

RADIA: RNA and DNA Integrated Analysis

RADIA identifies RNA and DNA variants in BAM files. RADIA is typically run on 3 BAM files consisting of the Normal DNA, Tumor DNA and Tumor RNA. If no RNA is available from the tumor, then it is run on the normal/tumor DNA pairs. For the normal DNA, RADIA outputs any differences compared to the reference which could be potential Germline mutations. For the tumor DNA, RADIA outputs any differences compared to the reference and the normal DNA which could be potential Somatic mutations. RADIA combines the tumor DNA and tumor RNA to augment the somatic mutation calls. It also uses the tumor RNA to identify potential RNA editing events.

The DNA Only Method (DOM) uses just the tumor/normal pairs of DNA (ignoring the RNA), while the Triple BAM Method (TBM) uses all three datasets from the same patient to detect somatic mutations. The mutations from the TBM are further categorized into 2 sub-groups: RNA Confirmation and RNA Rescue calls. RNA Confirmation calls are those that are made by both the DOM and the TBM due to the strong read support in both the DNA and RNA. RNA Rescue calls are those that had very little DNA support, hence not called by the DOM, but strong RNA support, and thus called by the TBM. RNA Rescue calls are typically missed by traditional methods that only interrogate the DNA.

PREREQUISITES

python (version 2.7)
The RADIA code is written and compiled in python. In order to run the commands, you'll need python (version 2.7).
samtools (tested on version 0.1.18 and 0.1.19)
RADIA uses samtools to examine pileups of reads across each sample in parallel.
You must install samtools prior to running RADIA.
pysam API (version 0.8.1 and higher)
RADIA uses the pysam API during the filtering process.
BLAT (optional)
RADIA uses BLAT to check the mapping of reads for all Triple BAM calls.
SnpEff (optional, tested on version 3.3 and 4.3)
RADIA uses SnpEff to annotate passing variants and to filter out calls from the Triple BAM method that land in genes with high sequence similarity.

DATA PREPARATION

BAM files
The BAM files need to be indexed with the samtools index command and located in the same directory as the BAM file itself.
FASTA files
The fasta files need to be indexed with the samtools faidx command and located in the same directory as the fasta file itself.

When running RADIA, you need to specify the appropriate fasta file - typically the one that was used during the alignment which is usually specified in the BAM header. Some fasta files use the "chr" prefix and some do not. Some BAM files use the "chr" prefix and some do not. If a BAM file uses the "chr" prefix, then the fasta file that is specified must also use the "chr" prefix and vice versa.

There are multiple ways to specify the fasta files when running RADIA. You can use the -f parameter for a fasta file that can be used for multiple BAM files. Typically, the fasta file for the normal and tumor DNA BAMs are the same. If the fasta file is not the same for all BAM files, you can overwrite the default fasta file specified with the -f parameter with the following BAM specific fasta files:

--dnaNormalFasta
--rnaNormalFasta
--dnaTumorFasta
--rnaTumorFasta

If the "chr" prefix is neeeded, then add the corresponding flag:

--dnaNormalUseChr
--rnaNormalUseChr
--dnaTumorUseChr
--rnaTumorUseChr

TEST SAMTOOLS COMMAND

In order to see if RADIA will be able to execute the samtools command without any errors, test the command prior to running RADIA. Here is an example command:

If the "chr" prefix should be used:

samtools mpileup -f fastaFilename.fa -E -r chr7:55248979-55249079 normalBamFilename.bam

If no "chr" prefix is needed:

samtools mpileup -f fastaFilename.fa -E -r 7:55248979-55249079 normalBamFilename.bam

RUN RADIA INITIAL COMMAND

RADIA is run on each chromosome separately. You need to specify the BAM files and the corresponding FASTA files. You can specify the output filename where the VCF files will be output, otherwise it will be sent to STDOUT. If the filename ends with ".gz", the VCF will be gzipped.

Run RADIA on 3 BAM files:

python radia.py
    patientId
    chrom
    -n normalDnaBamFilename.bam 
    -t tumorDnaBamFilename.bam 
    -r tumorRnaBamFilename.bam 
    -f hg19.fa 
    --rnaTumorUseChr 
    --rnaTumorFasta=hg19_w_chr_prefix.fa 
    -o /radia/raw/patientId_chr1.vcf.gz 
    -i hg19 
    -u http://url_to_fasta.fa

Run RADIA on 2 BAM files:

python radia.py 
    patientId
    chrom
    -n normalDnaBamFilename.bam
    -t tumorDnaBamFilename.bam
    -f hg19.fa
    -o /radia/raw/patientId_chr1.vcf.gz
    -i hg19
    -u http://url_to_fasta.fa

For the full list of optional parameters, type:

python radia.py -h

RUN RADIA FILTER COMMAND

There is one filter script that can generally be used for all filtering needs. Note: there is a difference between "flagging" and "filtering". A site that is "flagged" will have the information added to the INFO column of the VCF, a site that is "filtered" will have the filter added to the FILTER column of the VCF. By default all of the following filters are applied:

1000 Genomes Blacklists
Flag dbSNP
Flag Pseudogenes
Flag Retrogenes
Flag COSMIC sites
Flag RADAR sites
Flag DARNED sites
MPileup support
Read support

For calls that originate in the RNA, there are further filters:

Filter dbSNP
Filter Pseudogenes
Filter Retrogenes
Filter by BLAT (optional)
Annotate with SnpEff (optional)
Filter RNA genes and gene families

By default, all calls will be made on the data you provide.
If you only want the calls from the DNA Only Method, use the --dnaOnly flag
If you only want the calls from the Triple BAM method, use the --rnaOnly flag

If you want to exclude a particular filter, there are flags such as --noBlat to exclude the BLAT filter.

Many of the filters rely on data that is provided in the /radia/data/ directory. Other dependencies are on the pysam API and external programs such as BLAT (optional) and SnpEff (optional).

Here is an example filtering command:

python filterRadia.py
    patientId 
    chrom 
    /radia/raw/patientId_chr1.vcf 
    /radia/filtered/
    /radia/scripts/
    -b /radia/data/hg19/blacklists/1000Genomes/phase3/
    -d /radia/data/hg19/snp151/
    -r /radia/data/hg19/retroGenes/
    -p /radia/data/hg19/pseudoGenes/
    -c /radia/data/hg19/cosmic/
    -t /radia/data/hg19/gencode/basic/
    -a /radia/data/hg19/radar/
    -n /radia/data/hg19/darned/
    -s /snpEffDir/
    --rnaGeneBlckFile /radia/data/rnaGeneBlacklist.tab
    --rnaGeneFamilyBlckFile /radia/data/rnaGeneFamilyBlacklist.tab

Some default parameters to watch out for:

The default SnpEff genome is set to "GRCh37.75". If you are using a different version, be sure to upate the
--snpEffGenome parameter.
BLAT FASTA filename: by default the fasta file specified in the BAM header will be used. You can overwrite it with the
-f parameter. We recommend that you use a fasta file that includes the chrUn_gl000 contigs, chr_gl000_random contigs and the hap contigs (e.g. chr6_apd_hap1, chr6_cox_hap2, etc). Often times, the fasta files that are used during the alignment process of the bams exclude these contigs.
By default, the calls are filtered by the GENCODE basic gene regions. By specifying the --targetsInfo flag, the calls will be flagged (tagged in the INFO column) instead of filtered. If you don't want to flag or filter by target regions at all, use the --noTargets flag.

For the full list of optional parameters, type:

python filterRadia.py -h

RUN RADIA MERGE COMMAND

To merge all of the filtered chromosome files into one VCF file for the patient, execute the following command:

python mergeChroms.py
    patientId
    /radia/filtered/
    /radia/filtered/
    --gzip

This will merge all of the files with the names: patientId_chr*.vcf or patientId_chr*.vcf.gz into one file called patientId.vcf or patientId.vcf.gz (if you specify the --gzip parameter).

CITATION

If you use RADIA, please cite the method:
Radenbaugh AJ, Ma S, Ewing A, Stuart JM, Collisson EA, Zhu J, Haussler D. (2014) RADIA: RNA and DNA Integrated Analysis for Somatic Mutation Detection. PLoS ONE 9(11): e111516. doi:10.1371/journal.pone.0111516

LICENSE

RNA and DNA Integrated Analysis (RADIA) identifies RNA and DNA variants in NGS data.
Copyright (C) 2010 Amie J. Radenbaugh, Ph.D.

This program is free software: you can redistribute it and/or modify it under the terms of the GNU Affero General Public License as published by the Free Software Foundation, either version 3 of the License, or (at your option) any later version.

This program is distributed in the hope that it will be useful, but WITHOUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU Affero General Public License for more details.

A copy of the GNU Affero General Public License has been provided along with this program. Otherwise, see http://www.gnu.org/licenses/.

aradenbaugh/radia