When I was doing my first year of postdoc, I was supposed to extend my OPT (optional practical training), when I learnt that my university is no longer listed on e-verify and cannot hire me under a OPT any more. My OPT expires in about 40 days. I need to find another employer that is on e-verify before that in order to stay in the US. A friend very nicely referred me to a start-up company and I’ve got to meet the team. I remember two things from those interviews. The first one is this very professional lady, who worked for a prominent company, telling me that she joined this start-up due to family considerations. The other thing I remember was one of the questions from the CEO:”How do you distinguish rare variants from sequencing error?”

I don’t quite remember how I answered. In fact, to this day, I don’t know the answer.

Today let’s make some attempts at least try to understand the problem that we are facing.

Individual level vs. Population level

First of all, one need to understand that rare variants is a population-level concept, while sequencing error is at read level, and can be minimized at individual level.

Quality score

One obvious tool is quality score. Quality score of that base at fastq level, qulity score of the variant. If this variant got good coverage,