Whole Genome Sequencing (WGS) continues to provide informative insights into the genetic underpinnings of neurodegenerative diseases. In collaboration with the UCSF Memory and Aging Center and the University of Alabama at Birmingham for discovery cohorts and UCSB and the University of Antioquia in Colombia for a replication cohort, we conducted WGS for early onset AD and FTD cases. WGS was conducted for 657 cases and 927 controls in the discovery cohort, and an additional 540 samples and 152 controls from a Colombian cohort. An additional 7140 cases and 8237 controls were checked from ALS exome and genome studies, as well as ADSP (5722 cases and 5136 controls). Different filtering models were applied to the discovery cohorts for rarity (less than 1 in 10,000 or private), damage prediction (CADD 10 or 15 thresholds), and functional model (coding only or also including GenoSkylinePlus predicted regulatory regions). Rare variants in TET2 and PDZRN3 loci were enriched the discovery cohort after Bonferroni correction. Variants in PDZRN3 were almost all non-coding and were enriched for FTD over EOAD. Variants in TET2 were split between non-coding variants in putative strong regulatory regions and coding loss-of-function variants. Variants in TET2 were split approximately evenly between EOAD and FTD. In replication cohorts, Loss-of-function variants in TET2 were enriched in EOAD, FTD, and ALS, but not LOAD. The odds ratio for loss-of-function variants in TET2 vs. cohort controls for EOAD, FTD, and ALS was 4.8, and if population databases are brought in to the calculation, the odds ratio is 3.6. Data from discovery and replication cohorts suggest that loss-of-function variants in TET2 are associated with EOAD, FTD, and ALS, but not LOAD. The contribution from private non-coding variation emphasizes that this type of variation should be explored further. Further analysis of these cohorts, along with eventual meta-analysis with other similar cohorts of early onset neurodegeneration cases, is likely to provide important insights into the genetic underpinnings of early onset neurodegeneration.