The topic this quarter is Genetic Ancestry Inference. We will explore alternative strategies for estimating average ancestry proportions across the genome (Global Ancestry Inference, GAI) and at specific positions within the genome (Local Ancestry Inference, LAI). Each paper presents a novel method, usually providing theory, simulation, and an application.
This course is a journal club, with students presenting and leading a discussion each week. Students may also present their research in progress/capstone projects, with a separate reading and discussion board assignment for those weeks. Email Liz (em27@uw.edu) if you would like to schedule a presentation by Friday, October 6th.
Students will read each week's paper in advance, join the in-class discussion, and contribute to the online discussion board by noon each class day (Tuesday). The weekly discussion leader(s) will create a short presentation of the paper that provides an overview of the paper and facilitates further discussion.
Classes meet in-person for Autumn 2023, although we will have Zoom as a backup if the instructor or presenters cannot attend as per instructions across the UW. Class meets in T639 in the Health Sciences Building. It begins at 4:00 pm Tuesdays, starting with five minutes for announcements, followed by 45 minutes of discussion, ending at 4:50.
SPEAKER: Liz Blue
TOPIC: Introduction, review of resources, and speaker assignments.
We have covered similar topics before, see the reading lists for Autumn 2018 (Admixture), Autumn 2021 (Influence of ancestry on StatGen research) and Spring 2021 (Inferring history from DNA). You may also be interested in this infographic or this blog post describing how to read a scientific paper.
TOPIC: The original ADMIXTURE paper.
Fast model-based estimation of ancestry in unrelated individuals. by Alexander, Novembre, and Lange (2009).
This is a likelihood-based GAI approach, a classic tool used as a comparator in more recent papers.
TOPIC: The AIPS paper.
Ancestry inference using principal component analysis and spatial analysis: a distance-based analysis to account for population substructure by Byun et al. (2017).
This is a GAI tool that captures subpopulation structure within continents.
TOPIC: The GRAF-pop paper.
GRAF-pop: A Fast Distance-Based Method To Infer Subject Ancestry from Multiple Genotype Datasets Without Principal Components Analysis by Jin et
al. (2019).
This GAI tool is robust to small samples of markers and is used by dbGaP.
TOPIC: The RFMix paper.
RFMix: a discriminative modeling approach for rapid and robust local-ancestry inference by Maples et al. (2013).
This LAI tool implements a random forest approach and is widely used in applications and as a comparator in more recent papers.
Optional reading: The Ancestry_HMM paper.
A Hidden Markov Model Approach for Simultaneously Estimating Local Ancestry and Admixture Time Using Next Generation Sequence Data in Samples of Arbitrary Ploidy by Cobett-Detig and Nielsen (2017).
This hidden Markov approach (HMM) uses read counts rather than genotypes to perform LAI and estimate admixture time.
TOPIC: The Loter paper.
Loter: A Software Package to Infer Local Ancestry for a Wide Range of Species by Dias-Alves, Mairal, and Blum (2018).
This LAI approach uses a single-layer HMM.
TOPIC: The MOSAIC paper.
Fine-Scale Inference of Ancestry Segments Without Prior Knowledge of Admixing Groups by Salter-Townsend and Meyers (2019).
This two-layer HMM approach performs LAI, estimates Fst, and more.
TOPIC: The SALAI-Net paper.
SALAI-Net: species-agnostic local ancestry inference network by Sabat et al. (2022).
This is a neural network approach to LAI building on identity-by-descent.
TOPIC: The FLARE paper.
Fast, accurate local ancestry inference with FLARE by Browning, Waples, and Browning (2023).
This approach combines population genetics foundations with computational efficiencies to provide accurate LAI at scale.