Hi, I have low-depth genome sequencing data from several species in a family and I would like to build gene trees as input for analyses with programs like BUCKy and Astral. (I want to estimate a species tree, and assess evidence for introgression between nonsister lineages). I have mapped the reads to the closest outgroup available and produced VCF files for each scaffold.
I am wondering whether "gene trees" need to be real genes, or whether it is acceptable to simply cut the VCF file into small pieces and build "gene" trees of these pieces? It seems like most papers use real genes however I do not have a transcriptome for my organisms unless I used a fairly distant outgroup as a reference to assemble the gene sequences.