You are underselling our current ability to model the structure of DNA accurately in three dimensions based on sequence alone. With the advent of 3D-DART server in 2009
3D-DART: a DNA structure modelling server
DNA often changes its conformation as a result of interactions with various ligands; especially binding to proteins can…
and a ton of others before and since we are very good at doing this accurately. Also, while the structure of the DNA molecule is important in very specific situations, it has much less to do with its function then its sequence.
What we still struggle, though we have gotten much much better, to do is predict/calculate protein structure based on amino acid sequence alone.
I don’t know where you were intending to go with that from there.