The Open Genes Database of Associations with Aging and Longevity
The Open Genes database provides summaries of the information available on longevity-associated genes, from well-established and well-replicated effects such as that associated with klotho expression, to much less well supported data. Thousands of genes have at least a study or two suggesting an effect on longevity in studies of lower animals, and many of those may turn out to be experimental error. Yet since every fundamental cellular process can be influenced by dozens or hundreds of genes, even though there are comparatively few important processes of aging, one might well expect there be to be thousands of related genes. Looking at the practical outcome of all of this study, at the end of the day, the large human epidemiological databases and genetic studies suggest that common gene variants have very little effect on longevity when compared to the impact of lifestyle choices. Genetics, I suspect, is not the road to human rejuvenation. Instead, therapies that repair specific forms of damage will be needed.
The Open Genes database was created to enhance and simplify the search for potential aging therapy targets. We collected data on 2402 genes associated with aging and developed convenient tools for searching and comparing gene features. A comprehensive description of genes has been provided, including lifespan-extending interventions, age-related changes, longevity associations, gene evolution, associations with diseases and hallmarks of aging, and functions of gene products.
For each experiment, we presented the necessary structured data for evaluating the experiment's quality and interpreting the study's findings. Our goal was to stay objective and precise while connecting a particular gene to human aging. We distinguished six types of studies and 12 criteria for adding genes to our database. Genes were classified according to the confidence level of the link between the gene and aging. All the data collected in a database are provided both by an API and a user interface. The database is publicly available on a website at https://open-genes.org/.
So your hypothesis that genetic tweaking will not yield large results has some merit, as damage that has already been done may need to be literally repaired, and organs replaced if they are already in bad shape.
However, it is quite obvious that organisms are designed to fail over time, with molecular clocks being able to predict failure rates. It is entirely reasonable that certain cascades are meant to destroy the organism in favour of the group and it's evolution. This is of course passe in the age of intelligent evolution where an an organism that is already alive can be evolved.
So it could be that in order to prevent aging you may need to deactivate certain transcription factors and make other modifications. To assume that genetic modification will not make a huge difference is very wrong.
I wonder if anyone has tried to approach this as a big data AI problem? Do we have *enough* genetic data for people who have died of "age related diseases" (i.e., not car accidents) that one could train an AI on it? The idea would be to use the full genome as input, and the output would be a predicted age at death.
Perhaps the problem is that the human genome is too big for our current AI architectures? Perhaps it could be simplified to reduce the input size? The risk with this strategy is that you may accidentally pre-filter out something that it turns out is very important, like mutations in certain genes. Could we initially work with simpler organisms with a smaller genome to prove out the concept?