I have lots of (single) VCF files from a capture library which I notice are contaminated with what appear to be a lot of 8-oxoguanine artefacts (C>A and G>T) from the capture. Following from this paper https://www.ncbi.nlm.nih.gov/pubmed/23303777, I want to reduce the prevalence of these artefacts using a similar method, which means that I need to attach the trinucleotide context to each line of the vcf.

I can see plenty of packages which can summarise the context and signatures, but for filtering purposes it needs to be added to each entry so I can filter line-by-line. Is there a package/script which will allow me to manipulate the vcf directly to add these?