Closed mike-w-wilson closed 3 years ago
Meeting with Marcos Santoro and Elizabeth on July 16th to discuss his findings on optimal reference panel composition.
Marcos suggested the sample samples as above for the Native American component of the reference panel -- he recommended using the PEL samples that cluster closest to the other HGDP samples and drop the others. His work also showed that two and three way admixed samples can use the same reference panel.
Final reference is 241 samples: gs://gnomad-mwilson/lai/sample_selection/ref_to_run.tsv
pops for Native American component: 1kg: PEL
HGDP: Colombian Karitiana Maya Pima Surui