Updating and expanding inferred disease and phenotype annotations at Mouse Genome Informatics
Year: 2022
Type: other
Abstract: In the Mouse Genome Informatics (MGI, www.informatics.jax.org) database, phenotype and human disease annotations are created at the level of the genotype, our representation of the relevant allele and strain background of the mouse used in the studies. In order to provide the user with a summary of phenotypes and human diseases at the gene level, we developed an algorithm to infer gene-level annotations from the genotype annotations. The gene-level algorithm incorporates information about the full set of genome features associated with the alleles in a genotype, including endogenous and exogenous genes, to calculate the number of genes associated with the genotype. Additionally, tool alleles, such as reporters and recombinase expressing alleles, in the genotype are excluded prior to the final gene count. Recent improvements to the gene-level algorithm include expanding the criteria for identifying alleles that express additional genes and incorporating annotations of genes targeted by RNAi knockdown alleles. Inferred gene annotations are only created when the algorithm can identify a single gene associated with a genotype. As part of our work to integrate mouse data with other model organism data in the Alliance of Genome Resources (Alliance), we also create allele-level inferred annotations. To improve generation of allele-level annotations, we took advantage of enhancements to allele and genotype curation at MGI to develop an allele-specific algorithm. The allele-level algorithm uses many of the same rules as the gene level process to identify tool alleles. The primary difference between the two algorithms is that the allele-level rule set does not need to count the number of genes associated with a genotype. Thus, a genotype containing a single allele that mutates multiple genes, e.g. a multigenic deletion, may be used to create an inferred allele annotation but not an inferred gene annotation. However the allele-level algorithm does need to account for genotypes that involve multiple mutant alleles of the same locus. A genotype including two different mutant alleles of the same gene may be used to create an inferred gene annotation but not an inferred allele annotation. The inferred annotations generated by the updated gene and new allele algorithms are visible on the MGI and Alliance websites. Supported by NIH grant HG000330
Institution: Jackson Laboratory
Language: English
Subfield: Molecular Biology
Domain: Life Sciences
Open Access status: gold
OpenAlex ID: W4416509243