Statistical Genetics
Identifying human deafness genes with high density SNP chips
M Bahlo, C Bromhead in collaboration with K Smith (Murdoch Children’s Research Institute), M Hildebrand, R Smith (University of Iowa, IA USA)
In collaboration with the University of Iowa we have been identifying genomic loci responsible for deafness in twenty human pedigrees. The forms of inheritance vary considerably for these pedigrees and include autosomal penetrant, autosomal recessive and an apparent Y-chromosone linked deafness, which has only been reported once previously by another group. Most pedigrees appear to be harbouring fully penetrant mutations. These 20 pedigrees were represented by roughly 300 DNA samples. Each was genotyped with the Affymetrix 50K Xba SNP chip. Data was assembled and error checked with purpose written programs for this large collection of pedigrees.
We have been able to identify the genomic location of the mutation for six pedigrees and we have been able to identify three of the six mutations, with one mutation arising in a novel deafness gene. The other two variants are novel mutations in known genes.
We have encountered a false positive linkage signal owing to a single poor SNP chip sample. We have also identified a sample swap. In one pedigree we have observed several individuals who were deaf despite not carrying the genetic susceptibility haplotype. Re-examination identified that a sub branch of the family was inheriting a second, distinct form of deafness.
We have used a variety of linkage mapping algorithms including exact computation, combinations of exact computations and Monte Carlo Markov Chain Methods, depending on the size of the pedigree. We have also been experimenting with SNP marker set selection for the mapping to maximise mapping power.
Haplotype reconstruction for one deafness pedigree with only part of the pedigree shown. Squares represent men, circles women. Affected individuals are shaded in black. The two generations shown are products of a first cousin (second last generation) and a first cousin once removed marriage (last generation). This is an inbred pedigree with a recessive form of hearing loss. The blue genomic segment harbours the susceptibility locus. Affected individuals all have two copies of the blue haplotype in the critical region where the gene must be located.
Gene identification for pedigrees with hidden relatedness
M Bahlo in collaboration with D Amor, G McGillivray, E Fitzpatrick (Murdoch Children’s Research Institute)
With the availability of high throughput SNP data it is now possible to detect shorter regions of shared DNA than previously possible. This not only allows the mapping of recessive diseases caused by very distant relationships between parents (such as fourth or fifth cousins) but also allows the detection of such relationships. We are examining a variety of pedigrees where there are suspicions of relatedness. Detecting and testing for such relationships requires the determination of the background levels of inheritance by descent sharing for the appropriate population.
Genome wide association studies analysis
M Bahlo, M Moldovan in collaboration with J Stankovich (Menzies Research Institute, Hobart), S Browning, B Browning (University of Auckland, NZ), M Brown (University of Queensland)
Genome wide association studies are now being successfully used not just for complex diseases but also in pharmacogenomic approaches where exposure is too low to observe sufficient numbers of individuals in pedigrees. We are currently involved in the analysis of two such genome wide association scans. Each scan requires many checks to validate hits, as problems such as poor genotyping performance or undetected or uncontrolled experimental variation can cause false positives. Projects also require stringent multiple testing corrections. Confounding can also occur when generic controls are used or DNA was sampled from different sources.
Segregation analysis and mouse mapping using dense markers
M Bahlo in collaboration with G Davies (Immunology Division), J Silver (University of Copenhagen), SJ Foote (Menzies Research Institute, Hobart), H Dahl, S Manjii (Murdoch Children’s Research Institute)
ENU mutagenesis generates mouse mutants that are used to identify genes and genetic pathways in biological systems of interest. Subtle ENU mutants that show only small differences in phenoytpe in comparison to the wild type are difficult to verify as being heritable. We have been developing a statistical method that allows the estimation of genetic parameters, such as penetrances to infer appropriate genetic models. We have also been investigating newer, denser mouse mapping technologies such as the Affymetrix 5K mouse chip and the Sequenom mouse mapping panels and their applications to genetic mapping.
Navigation
Sponsors
