Mots-Clés
Palaeogenomics
Phenotype
Anthropobiology
ancient DNA
Description
From bones to biological traits - Developing a computational pipeline for phenotype prediction from ancient human DNA
We are offering a Master’s thesis project in the field of anthropobiology with a strong focus on bioinformatics. The project will be conducted at the Centre for Anthropobiology and Genomics of Toulouse (CAGT) under the supervision of Dr. Andaine Seguin-Orlando (PI of the ERC anthropYXX) and Anna-Lena Titze, a PhD student working in the AGES group, who is currently investigating patterns of kinship and social structure in Neolithic Southern France using archaeogenetic approaches. Anna-Lenas project is seeking to uncover patterns of relatedness at different scales of analysis (parent-child relationships, identity-by-descent sharing patterns linking individuals across centuries, population-level admixture) and explore how far burial practices reflect social organization, including potential selection criteria based on age, sex, or social status.
Background
Ancient DNA (aDNA) studies can provide direct insights into the genetic makeup of past populations, allowing researchers to explore human evolution, migration and population interactions over thousands of years. Phenotypic traits, such as pigmentation, lactase persistence, or immune-related characteristics, are key aspects of human biology that can be inferred from aDNA. Studying those traits from ancient populations helps us understand how human adapted to their environment, how selective pressures shaped our evolution, and how social and cultural factors influenced genetic variation.
The prediction of phenotypes allows to reconstruction aspects of appearance in prehistoric populations. For example, variations in skin, hair, and eye color can shed light on migration patterns, admixture events, and population continuity. Beyond appearance, functional traits such as disease sensitivity or dietary adaptations provide insights into the health, lifestyle and evolutionary pressures faced by ancient communities.
Working with aDNA, however, comes with unique challenges. Ancient samples are typically highly degraded. The preservation can vary widely depending on factors such as temperature, nature of the soil and storage conditions. One of the major challenges in ancient DNA research is modern contamination. This can be computationally detected, by instance, by exploiting predictable post-mortem damage patterns, including fragmentation and deamination at the end of ancient DNA strands. Another challenge lies in the possible evolutionary distances. Phenotypic predictions rely on modern reference datasets, which may not fully capture the allele frequencies or trait architectures present in ancient populations.
Project Description
Even though phenotypic traits are an important area of research to ancient DNA analyses, no comprehensive software currently integrates all necessary steps to obtain a phenotypic profile.
Therefore, the project aims to develop an automated pipeline that extracts phenotypic information from ancient genomic data and generates a clear, consolidated report in a single PDF.
Key objectives
• Identify genomic markers associated with phenotypic traits in humans
• Test methods on modern dataset with known genotypes
• Simulate ancient DNA from modern reference genomes using Gargammel
• Develop an automated analysis pipeline
• Generate clear output, including summary statistics and scientific figures
• Apply the pipeline to real ancient DNA data
• Presentation of scientific work
Requirements
• Enrolled in a Master’s program in Biology, Bioinformatics, Archaeogenetics or related fields
• Ability to communicate in English (spoken and written)
• Proactive and forward-thinking approach
• Prior programming experience (bash, R, and/or Python)