unlike linkage analysis whose goal is to identify candidate genes/variants based upon the association between polymorphic markers and the disease, I've found a candidate variant in a new gene through targeted NGS (gene pannel) that cosegragates with the phenotype in two families. How could I quantify the association between the variant and the disease? its very difficult to build haplotypes since I have discontinuous data (the sequenced genes are not homogeneously distributed in the genome).
In the field of candidate co-segregation, the number of bases sequenced is in the hundreds of thousands if not millions, and the number of family members for which the variant can co-segregate is measured in single-digits. The multiple-testing problem here is so significant that in practise, stats are never applied in published papers because they would be meaningless.
To add support for your novel variant, you should:
- Check in a general population of individuals suffering from the disease for variants (same or different) in the identified locus
- Knock it out/down in a mouse/fly/yeast and see what happens
- Isolate the potentially defective protein from the patient, and biochemically prove it is abnormal.
Of course this all depends on what the protein is, which past experience has taught me you should never tell anyone until after you have published -_-;
I totally agree with you. Thank you for your comments John