Recommendations & Troubleshooting

Haplotyping sliding frames with adjacent SNPs

In any situation in which neighboring SNPs are spaced apart within the length of a read, read-backed haplotyping can be used to phase SNPs. Here, we provide some recommendations for optimal parameter settings.

Use option -partial exclude

In case short regions of adjacent SNPs are haplotyped, only consider reads that span the entire locus. Otherwise, reads that only cover a part of the locus (by “random” shearing during library preparation and “random” read mapping start and stop positions) would create additional haplotypes marking absence of read coverage. For instance, a read could create a haplotype ‘000.’, if it was a reference allele of which the alignment stopped just before the last nucleotide to be haplotyped, and the “.” character denotes absence of read mapping. This haplotype is a technical artefact, not a biological signal.

Use option -mapping_orientation ignore

Because Shotgun reads may be mapped in any orientation (during Shotgun sequencing, genomic fragments are not cloned or sequenced with directionality with respect to the reference genome sequence), mode -mapping_orientation ignore should be used because then all reads are considered independent of their mapping orientation.

Use pair-aware read mapping

While the insert size of Shotgun libraries sequenced with Illumina instruments is relatively short (300-500 bp for paired-end libraries), paired-end reads (2x150 bp) usually do not overlap in the middle of the fragment and can not be merged during preprocessing. Read mapping should probably best be performed in pair-aware mode to increase specificity of mapping with BWA-MEM.

Less is more

Defining sliding frames in which to group adjacent SNPs is a trade-off between read depth, read length, and the density of SNPs. We recommend to create a set of BED files with varying sliding frame length and test these for locus and sample call completeness and correctness, and haplotype diversity (number of different haplotypes observed per locus across the sample set). As a rule of thumb, sliding frame length at about one-half to two-third of the read length provides an optimal balance between read depth and haplotype diversity and is a good starting point for further optimisation.

sliding frame lengthSNP densitycompletenesshaplotype diversity

The distance between the first and the last SNP within a maximal sliding frame length determine the effective sliding frame length. So, maximal sliding frame length may be optimised per sample set in function of the SNP density.

Haplotyping the junction sites of large structural variants such as deletions and inversions

Use option -partial include

The basic signal that is being detected is the localised and consistent lack of continued read alignment at a junction flanking a structural variant such as a (large-scale) deletion or inversion. So, reads are expected to show partial alignment in the three nucleotides that are covered in the sliding frame. In fact, only three haplotypes classes are commonly expected: 000 (reference); 00. ; 00- ; 0.. or 0– (upstream junctions) ..0 ; –0 ; .00 or -00 (downstream junctions).

Use option -mapping_orientation ignore

Because Shotgun reads may be mapped in any orientation (during Shotgun sequencing, genomic fragments are not cloned or sequenced directionally with respect to the reference genome sequence), mode -mapping_orientation ignore should be used because then all reads are considered independent of their mapping orientation.

Use single-end read mapping

While the insert size of Shotgun libraries sequenced with Illumina instruments is relatively short (300-500 bp for paired-end libraries), paired-end reads (2x150 bp) usually do not overlap in the middle of the fragment and can not be merged during preprocessing. Read mapping should probably best be performed as separate reads as large-scale rearrangements may cause large differences between the order of sequences in the reference and in the pair of reads. Thus, a larger number of reads may map onto the junctions, if each read can be placed independently of its paired read.

---

overviewrequired input

The scheme below shows how SMAP sliding-frames works downstream from variant calling and needs the VCF file with SNPs or SVs and the reference FASTA sequence as input.

---

Tabular output

BED file with sliding frames

By default, SMAP sliding-frames will return a BED file with the coordinates of sliding frames, used for SMAP haplotype-sites. The header below is only shown here for easy reference, it is not included in the actual output BED file.

Reference	Start	End	Locus_name	Mean_read_depth	Strand	SMAPs	Completeness	nr_SMAPs	Name
Chr1	99	200	Chr1:100-200_+	.	+	100,200	.	2	Frame_Set1
Chr1	449	600	Chr1:450-600_+	.	+	450,600	.	2	Frame_Set1