It looks like there are two new MIGPUMAs in the 2015 data set that are not in the 2010 PUMA crosswalk file. They are MIGPLAC=South Carolina and MIGPUMA=606, and MIGPLAC=Virgina and MIGPUMA=51001. Could you clarify which 2010 PUMAs are in those MIGPUMAS?
MIGPUMA1 606 in South Carolina and MIGPUMA1 51001 were originally grouped into the MIGPUMAs 600 and 51000 by the Census Bureau by mistake. This error was corrected for the 2015 ACS samples. In the 2015 ACS sample, STATEFIP 45, MIGPUMA1 606 is composed of PUMAs 601 and 602. STATEFIP 51, MIGPUMA1 51001 is composed of PUMAs 51089 and 51090. These codes should be reflected in the crosswalk after the next data release.