Help understanding biallelic marker requirement in Marker StatisticsCaveat: I have at best a high school-level understanding of genetics and may just be confused about something that's straightforward to an expert.
I just got off a call with a couple biology professors who wanted to know how to do Hardy-Weinberg Equilibrium tests. ("Simple!", I thought going in.) Then on the call they ask me specifically about this error message that they're getting from Marker Statistics:
Their data is from a diploid flower with a multiallelic marker (genotypes of AA, BB, CC, AB, AC, or BC, which they coded numerically 0 through 5), and according to the documentation, Marker Statistics can only handle biallelic markers (genotypes of AA, BB, or AB).
So a couple questions:
- Am I right in understanding that Marker Statistics can't handle their situation, and that this is the source of that error message?
- If the answer to 1 is "yes", is there a recommended response to a user who asks "Why can't Marker Statistics handle multiallelic markers?"
Specifically bugging @Chris_Kirchberg with this, but perhaps others might chime in too. Thanks.
Re: Help understanding biallelic marker requirement in Marker StatisticsHi @Ross_Metusalem,
It can only handle biallelic markers for a diploid. 0 homozygous major allele, 1 heterozygous, 2 homozygous minor allele. Major and minor are designations for frequency of occurance. I can ping Luciano Silva to see why we have limited it to balletic markers for diploids.
Re: Help understanding biallelic marker requirement in Marker StatisticsMarker Statistics allows only biallelic markers. Why? The implementation of formulas used for HWE, LD, and other measures in the platform are all biallic-specific.
A work around is to make biallelic markers out of the multiallelic ones. How? For each marker, take each allele one at the time and create a new biallelic marker for it. For example, say you have a marker with genotypes AA, AB, AC, BC, CC and DD. Start with allele A, go through each sample and call it 2 if it has two copies of A (AA), call it 1 if it has one copy of allele A (AB and AC) call it 1, otherwise (CC and DD) call it 0. Do the same for alleles B, C, and D. Notice, for each marker you will have as many biallelic markers as the number of alleles (in this example you will have four alleles, which will lead to four biallelic markers). Do the sample for all markers. In the end, you will have all biallelic markers which can them be used in Marker Statistics.
Re: Help understanding biallelic marker requirement in Marker StatisticsThanks, Luciano and Chris!
Luciano, your described workaround makes sense. I'll pass it along to the users.
