In one sentence
Computer simulations show how common genetic variation could tune shared neural-processing properties so some cognitive skills develop more easily than others — helping explain “spiky” profiles in autism, ADHD, dyslexia, and everyday individual differences.
What the researchers did
Heritable neurodevelopmental conditions often show uneven cognitive profiles (clear strengths beside clear weaknesses), yet genes rarely act only on one tiny adult brain “module.” The authors model a domain-relevance idea: genetic effects change substrate-wide computational properties; different skills then vary in how sensitive they are to those properties. They combine genetic algorithms with artificial neural networks over 20 simulated generations, selecting for performance on one of five tasks while tracking knock-on effects on the other four.
This is theoretical modelling, not a human genetics cohort or treatment trial.
What they found
- Selection that tunes shared processing properties for one task can either help or hinder other tasks, depending on each domain’s computational demands.
- Behavioural “deficits” in the model linked to higher heritability of individual differences.
- The work argues domain-relevance is a plausible route from polygenic DNA variation to uneven profiles — alongside (not necessarily instead of) more region-specific genetic effects.
- It offers a computational lens on why co-occurrence and heterogeneous profiles are common across dyslexia, ADHD, autism, language, and motor coordination conditions.
What this means for families and therapists
- Spiky profiles are expected, not proof of “laziness” or “not trying” in weaker domains.
- Support should build on strengths while scaffolding harder domains — without assuming one gene = one skill.
- Do not treat simulation results as personal genetic advice or a prediction for your child.
- Related reading: what we know about autism and practical school/support guides on the site.
Limitations and what we don't know yet
- Artificial networks are analogies, not real developing brains.
- Five abstract tasks simplify real cognition and socioemotional life.
- Does not specify which human alleles or interventions matter clinically.
- Needs empirical bridging studies linking genotypes, neural properties, and developmental trajectories.
This is a plain-language summary of The Genetic Origin of Uneven Cognitive Profiles in Heritable Neurodevelopmental Conditions and Individual Differences: Computational Investigations by Kohli M., Magoulas G., Thomas M.S.C., Developmental Science (2026). Source license: CC-BY. It is not medical advice — talk to a qualified clinician before changing assessment or therapy.