Bioinformatics Stack Exchange is a question and answer site for researchers, developers, students, teachers, and end users interested in bioinformatics. It only takes a minute to sign up.
Sign up to join this community
Why exactly does genotyping not maintain phase? I guess I'm more confused on what's the difference between sequencing and genotyping, and as a result how phase is maintained in one, but not the other. This is an image trying to do phasing using genotyping methods.
I understand how phasing works, I just don't quite get what is the difference between genotyping and sequencing, and why one of them doesn't work.
Phasing means that you are able to distinguish the both copies of your target sequence. (Usually you have one from your mother and one from your father).
If you can do this, depends mainly on the sequencing technology you are using. For example in case of sanger sequencing, you cannot do this, because you (pre)amplify all your copies at once and detect the sequence at the same time at the same location.
In case of ngs you have read information at the end. One read represents the sequence of one original dna molecule. If you detect multiple variants in one read, you can be sure that they are from the same chromosome and thus they are in phase. Now you can go on and try to find other reads that have unambiguous overlaps with these variants on your read. If you find some, you can also say that these variants have their origin on the same chromosome.
Many (small) variant calling algorithms use pileups to genotype variants position-by-position. This approach alone is insufficient to phase adjacent variants. However, if the variants of interest are close enough to be covered by a single read, it's certainly possible to phase them even with <100bp Illumina reads. In fact, I was implementing just such an approach recently when I asked this question about accessing bases by aligned position rather than by read position.
Genotyping (there are of course many different methods) is likely to give you information in the format of
Locus 1: 100% A
Locus 2: 50% C, 50% T
Locus 3: 50% C, 50% A
That's not phased.
Short or long read technology will give you data in the form of reads, each of which originates from one molecule, and if a read covers two loci, you'll know how they are phased. You will know if one DNA molecule has ACC, or ACA.
Sanger sequencing will give you mixed peaks at the locus of a mixed SNP, so there will be no phasing data.
