If a genotype is very rare we have a near violation of the positivity assumption. This can result in unfortunate high statistical confidence. The idea is to filter early those genotypes/queries with an arbitrary threshold. This will further reduce computational complexity.
If a genotype is very rare we have a near violation of the positivity assumption. This can result in unfortunate high statistical confidence. The idea is to filter early those genotypes/queries with an arbitrary threshold. This will further reduce computational complexity.