Alignment
This class takes a UCSC MAF file, a reference fasta, a sample name and an output file name. It creates a gvcf file from the MAF and reference, writing the data to the output file.
There are 6 optional boolean parameters:
fillGaps: Defaults to false. If true, and the maf file does not fully cover the reference genome, any gaps in coverage will be filled in with reference blocks. This is necessary if the resulting GVCFs are to be combined.
twoGvcfs: Defaults to false. If true, it indicates the input maf was created from a diploid alignment and should be used to create two separate gVCF files
outJustGT: Defaults to false. Output just the GT flag. If set to false(default) will output DP, AD and PL fields
outputType: Defaults to OUTPUT_TYPE.gvcf. Output GVCF typed files. If set to gvcf(default) it will send all REFBlocks, SNPs and Indels. If set to vcf, it will only output SNPs, Indels and missing values(if fillGaps = true)
delAsSymbolic: Defaults to false. If true, deletions larger than maxDeletionSize (see below) will be represented with the symbolic allele
.
There is 1 optional integer parameter:
maxDeletionSize: If delAsSymbolic is true, this is the maximum size of deletions that will be represented as simple deletions (not symbolic). This value does not include the padding base, and in the case where a deletion and insertion overlap, the difference between the ref and alt allele lengths is used. Must be positive. Defaults to 0 (deletions of any size are represented as symbolic)
These individual functions may be called: createGVCFfromMAF() - takes a MAF file, outputs a gvcf file. getVariantContextsfromMAF() - takes a MAF file, returns a list of htsjdk VariantContext records created from the MAF file data. exportVariantContext() - takes a list of htsjdk VariantContext records and exports them to a gvcf formatted file.
Requirements: Only 1 genome aligned to reference for this MAF file While MAF files may contain multiple records, for gvcf to MAF we need just 1. Multiple samples in the MAF record is much trickier processing as not every sample is necessarily in each alignment. This class is not handling that scenario.