Archetypes in human behavior and their brain correlates: An evolutionary trade-off approach

Abstract

Organisms perform multiple tasks and in doing so face critical trade-offs. According to Pareto optimality theory, such trade-offs lead to the evolution of phenotypes that are distributed in a portion of the trait-space resembling a polytope, whose vertices represent the specialists at one of the traits (archetypes).We applied this theory to the variability of cognitive and behavioral scores measured in 1206 individuals from the Human Connectome Project. Among all possible 300 combinations of pairs of traits, we found the best fit to Pareto optimality when individuals were plotted in the trait-space of time preferences for reward, evaluated with the Delay Discounting task. This task requires choosing either immediate smaller rewards or delayed larger rewards. Time preference for reward identified three archetypes, which accounted for variability on many cognitive, personality, and socio-economic status scores, differences in brain structure, as well as in functional connectivity between prefrontal cortex, basal ganglia, and amygdala, regions associated with reward and their regulation. There was only a weak association with genetics. In summary, time preference for reward reflects a core variable that biases human phenotypes via natural and cultural selection.

The architecture of functional lateralisation and its relationship to callosal connectivity in the human brain

“…The low dimensional structure of lateralised functions is consistent with previous findings that reported a low dimensional structure of functional networks 55 and of cognitive performance in both healthy controls 56,57 and patients 58 . Accordingly, individual performances or deficits are not task-specific but instead shared across a range of cognitive tasks.…”

Sperm length evolution in relation to body mass is shaped by multiple trade-offs in tetrapods

“…According to the Pareto Optimality Theory, archetypal species near the vertices should manifest maximum values in a set of traits and decrease in their scores monotonically by increasing the Euclidean distance from the vertex. We followed the approach used in 46 to analyze the feature density as a function of the Euclidean distance from the archetypes. For each feature we used the Wilcoxon rank sum test to test whether archetypes carried significantly higher values compared to the rest of the population 57 .…”

“…We investigated multiple, complex trade-offs between tasks using the novel statistical framework of Pareto Task Inference (ParTI) 34 . This methodology has been previously been employed in various biological systems, including gene expression 35 , tumours 36 and proteins 37,38 , trait morphology 34,[39][40][41][42] , neural systems 43,44 and behavioural traits 45,46 . We applied ParTI to investigate sperm size variation in relation to body mass in tetrapods, analysing an extensive dataset of 1388 tetrapods, which included 231 amphibians, 115 reptiles, 399 birds, and 643 mammals.…”