This problem has been addressed to some degree in prior tools like UCSC liftOver ( Fujita et al., 2011) and CrossMap ( Zhao et al., 2014). Variant-aware alignment methods require a step that translates, or ‘lifts over’ the alignments from the variant-aware references back to the standard reference coordinate system. While it is also useful to consider alternate assemblies, where the assemblies may lack a common coordinate system, we do not consider that scenario here. all references can be constructed from a linear reference together with a Variant Call Format (VCF) file describing differences from that reference. In these cases, variant-aware references use the original reference as a backbone i.e. A third approach is to use multiple population-specific references, which improves accuracy and reduces reference bias ( Chen et al., 2021). A second one is to specialize the reference using alleles specific to an individual ( Rozowsky et al., 2011). One such approach is to use the major allele reference ( Dewey et al., 2011), which substitutes common alternate alleles for rarer reference alleles. Some methods for reducing bias by substitute alternate alleles into one or more linear references in order to create variant-aware references. However, the use of a single linear reference leads to a phenomenon called reference bias-the tendency to produce incorrect alignments or entirely miss them for reads containing non-reference alleles. Most analyses of sequencing datasets start by aligning reads to a linear reference genome.
0 Comments
Leave a Reply. |