I want to map NCBI taxID, specifically a pair of Genbank sequences, to the divergence time estimate of their common ancestor. For example, using a coalescence theory approach predominantly used in phylogenetics. Do we have a database for such correlation?
I want to map NCBI taxID to their evolution time. Do we have a database for such correlation?
It would be best to edit your question to clarify what you mean by "evolution time" since this is not a well-defined term in molecular evolution. Even so, I think we can provide a clear answer to this question.
Estimating branch lengths between divergent taxa is an area of active research—there is no one method that works best in all scenarios. As far as taxonomy is concerned, resources like the NCBI Taxonomy database are under continuous refinement by the community.
Computing estimated divergence times and mapping those times to taxids would be a non-trivial task even for a single clade. Actively maintaining a database of divergence times across the entire taxonomy would be incredibly challenging, and it's doubtful that any such resource exists.
Frontiers in Ecology and Evolution did a focused and succinct review looking at 'the Tree of Life' in 2017 here. They particularly highlight the importance of integrating a fossil record within molecular trees. The upshot is simply that increasingly tighter integration between palaeontology and molecular dating is the ongoing research frontier.
The best approach in context is to look at relevant literature. The key challenge in my opinion is the underlying population dynamic because this does heavily affect temporal estimates.