Scientists are getting closer to completing a map of the mammalian immune system

Jul, 2021 - By WMR

The number and quantity of genes in the mouse and human genomes are remarkably similar, according to Beutler.

UT Southwestern researchers used artificial intelligence to find number of genetic changes in mice that are likely to damage their immune systems. The research is part of a Nobel Laureate's aim to discover virtually all such differences in mammals. The Center for the Genetics of Host Defense researchers identified 101 unique gene candidates with a more than 95% chance of essential life for immunity in this study, many of which they have already verified by re-creating mutations or knocking out genes.

They've also created Candidate Explorer, a piece of software that employs a machine-learning algorithm to detect chemically induced mutations that are likely to generate immune-related features. The software calculates the likelihood that any mutation they've caused would be confirmed as causing following additional testing.

The Nobel Prize in Physiology or Medicine was awarded to Beutler for his discovery of an essential class of receptors that allow mammals to promptly detect infection and generate an inflammatory response. The goal of Candidate Explorer is to aid researchers in determining whether a mutation linked to a phenotype trait or function is genuinely causal. Candidate Explorer has already aided in the discovery of hundreds of genes with novel immune activities. This will help us better understand the immune system so that we can identify new strategies to keep it healthy and understand why it occasionally fails.

If the researcher re-create the mutation out the gene, the candidate explorer give them a score that expresses them how possible it is that a particular mutation-phenotype association will be legalised for cause and effect.

CE looks into 67 aspects of the basic genetic mapping data to come up with a probability of causation estimate. Causation is obvious for certain mutations, but not for others. The algorithm "learns" over time through tests in which researchers re-create the mutation in a new pedigree and confirm or disprove the causal theory. This research, according to Beutler, is part of a nearly ten-year effort to find every mutation that could damage the mouse immune system.