The Ecology of Genes

I can’t stop thinking about the really remarkable discoveries about imprint genes made last week by my Duke colleague Alex Hartemink, an expert in bioinformatics (computational science and biology).  Here’s the url for the news release (“Duke Scientists Map Imprinted Genes in Human Genome): http://www.dukemednews.duke.edu/news/article.php?id=.   So much popular press about genetics in the last decade has been deterministic and mechanistic.  “Cancer families” is one phrase I’ve heard too many times, as if one’s family tree decrees one’s immutable, malignant fate.   But if this finding about at least 156 imprinted genes is correct, it may be that what one inherits is not genes that determine our fate but genes that are mutable, subject to environmental, dietary, emotional, and pharmacological change during the course of the life of individuals and populations.  As we find out more about imprinted genes, it may turn out that the world we create (on all levels) is the most important factor in determining the diseases to which we are susceptible. What moral and social rules should we uphold if we are not genetically determined but genetically responsible for our fate and that of others?  This is very preliminary, of course, but it is paradigm-turning to think about the imprinted genes now implicated in cancer, diabetes, autism, obesity, and several other major illnesses and conditions.  And I am very excited at the role of computation in being able to decode what was previously inaccessible.  To make this discovery about imprinted genes, Alex’s team used the form of Artificial Intelligence known as Machine Learning which is based on algorithms which allow computers to “learn” from their previous mistakes and even to analyze biases in their own past programs through inductive principles based on Bayesian statistics (a method based on collecting evidence from contradictory hypotheses to infer the probability that one or another inference is more likely to be true). At breakfast this morning with my friend Bob who is visiting from out of town, we talked about how much we miss the voice and the writing of the late Stephen Jay Gould. I would have loved to read  Gould’s spinning of the implications and possibilities for a new ecology of (imprinted) genomics.