Have a look at vg convert here:

https://github.com/vgteam/vg

These are mostly valid examples or at least they were since there are a lot of formats in vg.

pangenome_mapping.nf:    vg convert -t $task.cpus --gfa-in $gfa -p > ${prefix}.vg
pangenome_mapping.nf:    vg convert -t $task.cpus -x -g ${gfa}.fixed.gfa > ${gfa}.xg
pangenome_mapping.nf: *  convert gam to gaf
pangenome_mapping.nf:    vg convert --threads $task.cpus $vg --gam-to-gaf $gam > ${prefix}.gaf 
pangenome_mapping.nf: *  convert gaf to gam
pangenome_mapping.nf:    // Should be: vg convert graph.vg -G alignment.gam > alignment.gaf
pangenome_mapping.nf:    vg convert --threads $task.cpus $vg --gaf-to-gam $gaf  > ${prefix}.gam  
runbatch_vg_index.sh:   $vg_bin convert -x -g $i.fixed.gfa > $i.xg
runbatch_vg_index.sh:#vg convert -x -b graph.gbwt graph.gg > graph.xg
runbatch_vg_index.sh:#vg convert -x -b graph.gbwt --ref-sample ref graph.gg > graph.xg