E-Locus - Ιδρυματικό Καταθετήριο Πανεπιστημίου Κρήτης - Inference of population relations using ancient genomes : challenges, biases and imputation approaches in dimensionality reduction methodologies

Αρχική Inference of population relations using ancient genomes : challenges, biases and imputation approaches in dimensionality reduction methodologies

Αποτελέσματα - Λεπτομέρειες

[Προσθήκη στο καλάθι]

Κωδικός Πόρου

000448414

Τίτλος

Inference of population relations using ancient genomes : challenges, biases and imputation approaches in dimensionality reduction methodologies

Άλλος τίτλος

Εξαγωγή συμπερασμάτων πληθυσμιακών σχέσεων χρησιμοποιώντας αρχαία γονιδιώματα: προκλήσεις, μεροληψίες, και προσεγγίσεις για ανάκτηση της ελλιπούς πληροφορίας σε μεθόδους μείωσης διαστάσεων

Συγγραφέας

Παπαδοπούλου, Αγγελική Β.

Σύμβουλος διατριβής

Παυλίδης, Παύλος

Μέλος κριτικής επιτροπής

Ηλιόπουλος, Ιωάννης
Νικολάου, Χριστόφορος

Περίληψη

Ancient DNA (aDNA), derived mainly from archaeological findings, is a snapshot into the past, enabling researchers to obtain insights into the human evolutionary history. Due to its age, aDNA is highly degraded and damaged, leading to limited, often low-quality, extracted information. aDNA studies unravel the complex evolutionary history of human demography, such as the past population size changes, gene flow events as well as adaptive processes that contributed to the survival of our species. The demographic inference of ancient populations sheds light on the genetic relationships between them assisting us to understand the present-day structure and the common origins of human populations. In this study, we describe and evaluate widely-used methods for the inference of population structure, focusing on the Dimensionality Reduction techniques. We highlight the biases, introduced by the inevitable missing data in aDNA. We propose a novel imputation approach that is based on the phylogeny of the samples under study. Using simulations, we tested the accuracy of our imputation approach and we showed that is superior to the widely-used mean imputation and has a similar performance to the state-of-the-art kNN imputation. In conclusion, this thesis draws attention to the challenges accompanying the usage and analysis of aDNA data to infer population relationships and, in addition, proposes a novel imputation approach to retrieve the missing information of aDNA genotypes.

Φυσική περιγραφή

iv, 80 σ. : πίν., σχήμ., εικ. (μερ. εγχρ.) ; 30 εκ.

Γλώσσα

Αγγλικά

Θέμα

Ancient DNA

PCA

Population structure

Πληθυσμιακή δομή

Ημερομηνία έκδοσης

2022-07-29

Συλλογή