divergence dating under partitioned single-gene model

[bryans...@gmail.com]

I am working on dating a large single-species phylogeny (190 samples) based on a single mitochondrial gene. I have an estimated mutation rate/clock rate from the literature. The BEAST2 analysis (and resulting log/trees) look reasonable. However, I am struggling with how to implement the clockRate when partitioning the data by nucleotide position. I do not have rates for every position, just a whole-gene rate. 

Is it still possible to use a clockRate applied to the entire gene while using site models for partitions? If no, would appreciate hearing strategies to use relative rates in a correct way. 

Thanks in advance!

Bryan