ATLAS / FUNCTIONAL TRAITS

DNA can explain some variation. It does not choose a person’s path.

We review exercise, cognition, sleep and metabolism—separately from clinically actionable genetic conditions.

THE CORE LIMIT

A functional trait ≠ a talent test

There is no validated genetic test that reliably predicts an individual’s outcome in sport, intelligence, personality or career. The more complex the trait, the more development, environment, learning, health and observed function matter.

01 / TRAIT MAP

What people ask—and what can actually be inferred.

01Exercise and sport

Athletic ability

ACTN3 · ACE · PPARGC1A · thousands of variants

Question
Can DNA choose a sport for a child or adult?
Evidence
The trait is heritable, but individual variants explain too little of performance.
More useful
Observe actual skills, interest, health, training response and recovery.
Limit
Few-SNP panels are not validated for talent selection or training prescription, especially in children.
02Exercise and sport

Response to exercise

Polygenic · muscle · cardiovascular · behavioural factors

Question
Does a genome predict gains in strength, endurance or injury risk?
Evidence
Group-level associations reproduce unevenly; environment and training design contribute strongly.
More useful
Functional assessment, history, movement technique and load tracking provide direct information for planning; consumer DNA panels have no validated individual prediction.
Limit
Do not use COL5A1, ACE or ACTN3 alone to predict injury or performance.
03Cognition

Learning and cognitive ability

Highly polygenic · development · education · environment

Question
Can genes determine intelligence, memory or an educational path?
Evidence
Population GWAS find many tiny effects; individual prediction is constrained by ancestry and environment.
More useful
For a child, use development, learning, hearing, vision, sleep and clinical assessment when concerns arise.
Limit
A polygenic score does not measure talent and must not determine schooling, career or family expectations.
04Cognition

Personality and behaviour

No single personality gene

Question
Is there a ‘gene for anxiety’, leadership or risk taking?
Evidence
Personality traits are polygenic; individual variant effects are minuscule.
More useful
Use validated psychological methods for their intended purpose, not genetic labels.
Limit
DRD4, COMT and SLC6A4 do not produce a reliable personal profile.
05Sleep and rhythms

Chronotype and sleep

PER3 · CRY1 · CLOCK pathways · polygenic background

Question
Can DNA tell whether you are a morning or evening person?
Evidence
Large studies identify hundreds of loci, but explain only part of the variation.
More useful
Sleep diaries and observed functioning describe the actual pattern; a few SNPs add little to an individual decision.
Limit
The result does not diagnose insomnia, apnoea or a circadian disorder.
06Nutrition and metabolism

Lactose tolerance

MCM6 regulatory variants → LCT

Question
Does a variant indicate adult lactase persistence?
Evidence
The relationship is strong, but causal markers differ across populations.
More useful
Interpretation is combined with symptoms and clinical assessment when needed.
Limit
European rs4988235 is not a universal test and does not prove the cause of symptoms.
07Nutrition and metabolism

Caffeine and stimulants

CYP1A2 · ADORA2A · many non-genetic factors

Question
Can genotype calculate a safe caffeine dose?
Evidence
Metabolism and sensitivity associations exist, but no clinical personal-dose model is established.
More useful
Consider age, pregnancy, medicines, sleep, blood pressure and observed response.
Limit
It is not suitable for prescribing caffeine to children or treating fatigue.
08Clinically actionable

RYR1 and serious exercise/anaesthesia reactions

RYR1 · CACNA1S

Question
When is exercise genetics clinically actionable?
Evidence
Pathogenic variants are linked to malignant hyperthermia and some rhabdomyolysis episodes.
More useful
Testing is considered after a personal or family episode or suspicious anaesthetic/exertional reaction, through a specialist.
Limit
This is not a population sports test; a negative limited panel does not eliminate risk.

02 / CHILDREN

Not ‘every DNA test a child needs’, but tests with a clear benefit for the child.

Sports, intelligence, nutrition or personality panels are not necessary screening for a healthy child. Testing is justified when it answers a clinical question and can change care during childhood.

01

Newborn screening

When: For all newborns under the national programme in the relevant country.

How: Usually biochemical and physiological screening; DNA testing confirms selected findings.

Why: Early detection of conditions where timely action changes outcomes.

02

Diagnostic testing

When: For congenital anomalies, developmental delay, intellectual disability, seizures or another clinical suspicion.

How: Karyotype/CMA, targeted panel, exome or genome—selected for the phenotype.

Why: Can refine diagnosis, surveillance, treatment and family risk.

03

Familial variant actionable in childhood

When: When a pathogenic familial variant is known and surveillance or prevention begins before age 18.

How: Exact targeted confirmation of the familial variant after genetic counselling.

Why: An informative positive or true-negative familial result.

04

Medication-linked pharmacogenetics

When: When a child is receiving a specific medicine with an applicable guideline or label.

How: Validated testing of the relevant gene/allele, not a universal wellness panel.

Why: A clinician may refine choice or starting dose alongside clinical monitoring.

DO NOT USE AS A DECISION BASIS

Sports selection, educational restriction, intelligence prediction, career choice, symptom-free diets or personality labels based on a few DNA variants.

03 / PRIORITIES: VALUE × RISK

High value never removes the clinical gate.

This register contains the high- and very-high-value tracks. Risk is assessed separately: the closer a feature comes to diagnosis, medication, children or reproductive decisions, the stricter its sources, stop rules and professional review.

01
Very high valueControlled risk

Local genetic-variant reference

Use
Checks a standardised designation and displays an exact record from a limited ClinVar snapshot without uploading a VCF or report.
Output
Source record, classification, review status, date and inference boundary.
Gate
Exact match; the query stays in browser memory; no match is not interpreted.
02
Very high valueMaximum clinical risk

Oncology clinical-evidence navigator

Use
Shows when a biomarker and clinical context match an exact regulatory source—and when context is insufficient.
Output
Not a prescription: a verifiable relationship for discussion with an oncologist and molecular tumour board.
Gate
Structured fields, exact indication, jurisdiction and source date only; no doses, ranking or VUS output.
03
Very high valueMaximum clinical risk

Medication-safety pharmacogenetics

Use
Matches an exact gene–medicine pair to a relevant professional guideline or official regulatory label without merging their statuses.
Output
Evidence status, required clarifications and clinician questions—not a personal dose.
Gate
Confirmed genotype/phenotype, exact medicine and clinical review; complex cases are blocked.
04
Very high valueHigh clinical risk

Inherited cardiometabolic pathways

Use
Separates familial hypercholesterolaemia, cardiomyopathies, arrhythmias and monogenic diabetes from weak wellness-risk claims.
Output
A pathway for checking the report, clinical phenotype and questions for family counselling.
Gate
Exact gene–condition validity, pathogenicity, phenotype and clinical assessment—not a generic polygenic score.
05
High valueControlled risk

Test audit and source freshness

Use
Helps evaluate intended use, method, coverage, limitations, privacy and clinical utility before purchase.
Output
A multi-axis transparency profile without a misleading universal quality score.
Gate
Versioned primary sources, review date and independent conflict disclosure.
06
High valueHigh clinical risk

Rare-disease diagnostic pathway

Use
Explains when a panel, CMA, exome, genome or familial-variant test fits the clinical question.
Output
A map of method, possible results, limitations and next step with a genetics professional.
Gate
No diagnosis from a web form; phenotyping and method selection remain clinician-led.
07
High valueMaximum clinical risk

Carrier, reproductive and newborn pathways

Use
Separates couple carrier screening, prenatal screening, diagnosis and national newborn screening.
Output
Purpose, residual-risk, confirmation and genetic-counselling explanation.
Gate
Separate child and reproductive ethics; no automated recommendation.
08
High valueHigh clinical risk

GDC tumour-cohort research context

Use
Shows alteration frequency and research context in de-identified cancer cohorts.
Output
Aggregate research visualisation kept separate from clinical actionability.
Gate
Never used for individual treatment selection, prognosis or bypassing controlled-access data.

04 / IN DEVELOPMENT

A living register of the next modules—from clinical pathways to research boundaries.

This is an editorial roadmap, not a promise of personal prediction. Before publication, every topic receives an exact relationship, measurement method, evidence, clinical actionability and an inference boundary.

01Question-dependent

Appearance and senses

  • Skin, hair and iris pigmentation
  • Hair texture and androgenetic hair loss
  • Freckles, sun sensitivity and phototype
  • Taste, smell, colour vision and selected hearing phenotypes

Common appearance traits stay separate from monogenic vision, hearing or pigmentation disorders.

02Question-dependent

Sleep, rhythms and the nervous system

  • Chronotype, sleep duration and regularity
  • Rare familial circadian-rhythm disorders
  • Restless legs and narcolepsy—only in clinical context
  • Migraine, pain and sensory sensitivity

A polygenic signal does not replace a sleep diary, examination or diagnostic testing.

03Question-dependent

Exercise, recovery and sports safety

  • Aerobic adaptation, strength and muscle-fibre composition
  • Bone, tendon, ligament and injury traits
  • Recovery, inflammatory response and thermoregulation
  • Rhabdomyolysis, cardiomyopathies and anaesthetic risk

SNPs do not determine talent or a training plan; clinical red flags receive a separate pathway.

04Question-dependent

Nutrition and metabolism

  • Lactase, caffeine, alcohol and taste receptors
  • Lipids, glucose, uric acid and iron
  • Coeliac disease: HLA as an exclusion aid, not a diagnostic test
  • Vitamins and nutrients without ‘DNA diet’ promises

Genotype does not calculate a universal diet or a safe substance dose.

05Clinical pathway

Cardiometabolic traits

  • Familial hypercholesterolaemia and Lp(a)
  • Inherited cardiomyopathies and arrhythmias
  • Monogenic diabetes and rare metabolic disorders
  • Polygenic risks for blood pressure, diabetes and body weight

Monogenic diagnoses, measurable biomarkers and polygenic risks are shown separately.

06Question-dependent

Immunity and inflammation

  • HLA relationships with autoimmune conditions
  • Pharmacogenetic HLA alleles
  • Rare primary immunodeficiencies
  • Allergy and infection susceptibility—without consumer promises

An HLA risk association is not a diagnosis or a general ‘immunity test’.

07Clinical pathway

Reproductive health and life course

  • Carrier screening and couple-level risk
  • Genetic causes of infertility and premature ovarian insufficiency
  • Prenatal screening and diagnostic testing
  • Newborn screening and early diagnosis

Screening, diagnosis and prediction for a healthy person must not be conflated.

08Not for personal prediction

Cognition and behaviour

  • Memory, attention and educational outcomes
  • Personality, risk-taking and addictions
  • Neurodevelopment in the presence of clinical symptoms
  • Neurodegenerative risks and the right not to know

We do not create talent, personality or schooling scores, especially for children.

09Clinical pathway

Medicines and adverse reactions

  • CPIC/DPWG relationships for exact gene–drug pairs
  • Oncology toxicity: DPYD, TPMT, NUDT15 and UGT1A1
  • Antiplatelets, statins, immune reactions and anaesthesia
  • Limits of combinatorial psychiatric PGx panels

Every recommendation is tied to a medicine, indication, guideline and clinician-led decision.

10Research zone

Ageing and lifespan

  • APOE, FOXO3 and population longevity associations
  • Cardiovascular and cancer contributors to long-term risk
  • Clonal haematopoiesis as an acquired, not inherited, signal
  • Epigenetic clocks as a separate biomarker, not a genomic forecast

A group-level association is never converted into a promise of extra years.

This register is not an exhaustive list of human traits and will expand only where a verifiable source and clear use boundary exist.

05 / NEXT STEP

Question first. Method second. Genetic test third.

Audit a productRead the traits review