A Constellation Model of Biology
We are all made of stars
BIOLOGYEPIGENETICSANCESTRY
Alexandra Chambers
8/23/20263 min read
What I have been building for the past year with my Divergent Genomics framework is, structurally, a Constellation Model of Biology. My framework has repeatedly arrived at the same principle from different directions: the biologically meaningful unit is often not the individual variant, diagnosis, anatomical feature or biochemical pathway - it is the configuration.
This work keeps refusing linear categories. 'Autism' alone tells you very little. 'EDS' by itself tells you very little. An unusual uterus, skeletal morphology, immune reactivity, methylation architecture, sensory phenotype, mitochondrial metabolism, HLA pattern, circadian differences, blood group, craniofacial development or an archaic allele considered separately can all look like disconnected observations. However, when particular features recur across functional systems and across people, the pattern between the points becomes the information. That is a constellation. For example, total Neanderthal DNA percentage isn't necessarily the important variable; where the inherited material sits, what regulatory architecture surrounds it, what other variants it is interacting with, and which developmental systems are consequently altered is more important than the raw quantity. Two people can therefore carry overlapping components but produce very different phenotypes because the overall arrangements differ.
There is no single “divergent genome.” You get divergent genomic constellations. One constellation might be weighted towards connective tissue, skeletal morphology, autonomic regulation and sensory processing. Another towards immunity, metabolism, methylation and neurodevelopment. Another could involve reproductive architecture, extracellular matrix, craniofacial development and immune compatibility. There are shared stars, but different configurations. There's also another layer that I think is particularly relevant - a constellation does not actually exist as a line drawn in the sky. The stars exist. The boundary and the name are the observer's attempt to recognise a recurring configuration.
That is the epistemological problem I have been scrutinizing. “Autism.” “EDS.” “MCAS.” “ADHD.” “Fibromyalgia.” “FND.” “PCOS.” Medicine draws circles around particular collections of observable phenomena and gives each collection a name. Divergent Genomics asks what happens when you stop staring at the circles and start looking at the entire sky. Some supposedly separate diagnoses then begin sharing stars. Some stars turn out to belong to processes operating across several systems: extracellular matrix biology, developmental signalling, folate/methylation metabolism, immune regulation, mitochondrial function, neural crest development, endocrine signalling, mechanotransduction, bioelectric signalling, epigenetic regulation. That is why my research keeps migrating across disciplinary boundaries. I am following the connections rather than maintaining the artificial borders drawn around them.
The skeletal-system article I am working on is another expression of exactly this idea: bone, teeth, palate, skull, spine, pelvis and connective tissue aren't independent objects. They're surviving manifestations of developmental processes involving timing, folding, fusion, signalling, mechanical force and remodelling. The adult structure is the visible constellation left by development.
Then there is the evolutionary level. Functional hybridisation has never really been about somebody being 4% Neanderthal, it's about the constellation of introgression. I've been investigating non-random arrangements of evolutionarily informative variation across interacting functional systems - immune, sensory, connective-tissue, metabolic, circadian, neurodevelopmental and regulatory architecture.
Biologically, we're quite literally assembled from elements forged in earlier generations of stars - carbon, oxygen, calcium, iron and so forth. So stars themselves are not entirely metaphorical either. Stellar nucleosynthesis produced the material - evolution configured it - development organised it - environment continually modifies its expression.
Stars that appear beside one another in a constellation can actually be separated by enormous distances. Seeing a pattern therefore doesn't automatically establish a causal connection between every point. It still has to be established which relationships are mechanistically valid, which are developmental, which are inherited, and which merely co-occur. That distinction is mechanistically what prevents constellation thinking from becoming pattern-seeking without constraint.
Humans are the visible expression of a genomic, developmental and environmental constellation.
I will be documenting a proposed Constellation Model of Biology in a repository and I will link it here when I have done so.
Image: Flammarion engraving, unknown original artist, first published in Camille Flammarion, L’Atmosphère: Météorologie Populaire (1888). Colourisation by Hugo Heikenwaelder, Vienna, 1998. CC BY-SA 2.5 / Wikimedia Commons.