Hi Paula,
From the team that handles these activities now and for the Connectivity Atlas:
The ML/AP/DV coordinates for all of the injection targets were initially obtained using a Paxinos atlas referenced to bregma. However, as noted, we used the surface of the brain at those M/L and A/P coordinates, as the zero point for the DV coordinates. Importantly, we did make adjustments to coordinates depending on feedback from our histology team regarding targeting accuracy. Additionally, it’s important to note that all of the injections for the connectivity atlas are then translated into the CCF, which unfortunately does not register back to stereotaxic coordinates. As for superficial cortical injections, the user is correct that it’s best to angle the injector such that it is normal to the laminar structure of the cortex at the M/L and A/P coordinates. As coordinates moved laterally, we went from 0 to 20 to 30 degrees to maintain normality as best we could.
Lastly, here is a link to our technical white paper: Documentation: Mouse Connectivity Atlas