Does the CN with LQ need to be deleted?
Is the CN with LQ in the svgenotype.vcf file reliable? Does it need to be deleted in subsequent analysis? Such as calculating the coefficient of variation .Or after the default parameters are filtered, these LQs can be ignored, and CN is still reliable.
Thank you for your post. Bob Handsaker has been tagged and will get back to you shortly.
I'm assuming you mean genotypes with "LQ" as the value of the FT tag in the genotype field. The LQ filter value is simply a shorthand for CNQ < 13 (or roughly 95% confidence) in the diploid copy number estimate. Whether you want to use these genotypes or not depends on what you want to do.
For the coefficient of variation (of the true copy number), my reaction without thinking about it too hard is that dropping all LQ genotypes would probably lead to a biased estimate since DNA samples with higher copy number will tend to have less confident genotypes, which will bias downward the estimate of both the standard deviation and the mean. I.e. it is harder to tell CN 9 from CN 8 than it is to tell CN 2 from CN 3 or CN 1.
Please sign in to leave a comment.