Inferring human phenotypes using ancient DNA: from molecules to populations.

Area of Science:

• Paleogenomics
• Computational Biology
• Human Evolution

Background:

• Ancient DNA (aDNA) analysis has advanced understanding of ancient human population movements.
• Inferring ancient phenotypes from aDNA is limited by the genotype-to-phenotype mapping complexity.

Purpose of the Study:

• To review recent advancements in inferring ancient phenotypes using molecular and population-level data.
• To explore the application of machine learning (ML) approaches in ancient phenotype prediction.

Main Methods:

• Review of current literature on ancient phenotype inference.
• Focus on machine learning algorithms applied to aDNA data.
• Integration of molecular phenotype predictions with non-genetic data.

Main Results:

• Advances in inferring ancient phenotypes from molecular to population scales are highlighted.
• Machine learning approaches offer new possibilities for phenotype prediction from aDNA.
• Predicting complex individual traits across evolutionary distances remains difficult.

Conclusions:

• Integrating conserved molecular phenotype predictions with non-genetic data offers a viable strategy for understanding ancient human traits.
• Future research should focus on refining ML models and data integration for robust ancient phenotype reconstruction.