In class, the professor said that there is about 8% of protein-coding genes in human. I wonder how biologists estimate this without annotating all of human genes.