Dr Kate Elliott

Statistical Geneticist/Bioinformatician

Kate Elliott

Research summary

GeneSniffer

I have developed a database mining program, GeneSniffer (genesniffer.org), which assigns a "disease candidacy" score to genes. For each gene the method gathers literature from PubMed and information pages (eg from Entrez and OMIM, and mouse orthologs from Jackson) and searches these with a list of diseases specific keywords to generate a candidacy score. Information is also added from homologous and interacting genes and added to the score. The method can be used for any disease or trait of interest across any list of genes (an unordered list, surrounding signal SNPs, chromosomal regions or the whole genome). Definition of regions surrounding SNPs of interest is very important since LD patterns will predict the spatial range of the functional variant giving rise to a signal. GeneSniffer has been adapted to take SNPs as an input and analyses any genes within the LD region of the SNPs. GeneSniffer can be rerun at any time to incorporate database updates and saves a huge amount of time compared to sifting through the databases manually looking for interesting candidates.

Genetic overlap between diseases from GWA data comparisons

In addition I have been developing techniques to look at the genetic overlap between different diseases using GWA data. These methods include looking at the distribution of concordant and discordant risk alleles between disease, comparing the number of risk alleles for disease 1 in cases and controls for disease 2, LASSO (testing whether disease 1 can predicts case control status in disease 2) and looking at the distribution of disease 1 signals at proven signals or genes for disease 2 and vice versa. These methods have been applied to comparing T2D and cancer datasets.

Publications - see under listing

Research Areas

Gene candidacy Text mining Type 2 diabetes Cancer

Keywords

Genesniffer Text-mining Type 2 diabetes Cancer GWA-overlap Linkage disequilibrium