arXiv subjects

Carlos D. Bustamante

Publications and source records attributed to Carlos D. Bustamante.

11 recordsLinked to original sources

arXiv2026

The Human Genomic Landscape of Oceania

Oceania and Island Southeast Asia have a rich, yet understudied, human genomic landscape. This region encompasses some of the first areas inhabited by humans following the out-of-Africa expansion, includes populations with the highest levels of archaic hominin introgression, and contains Pacific islands that are among the most remote continuously inhabited locations in the world. Here, we describe the first region-wide analysis of individuals from population groups spanning Oceania and its broad perimeter. In total we generate and analyze genome-wide data from 92 different populations, 58 separate islands, and 30 countries, covering one third of the planet. Leveraging this diverse dataset, we resolve genetic connections among islands, providing a detailed view of regional population structure and identifying the island groups involved in the settlement of several Polynesian Outliers. Ancestry-specific analyses allow us to deconvolve different layers of history, from tracing groups deriving their Austronesian ancestry via the Lapita expansion to quantifying variable archaic introgression across the basal Papuan component of Oceanians and Southeast Asians. Finally, we map biomedically relevant variants across Oceania and Southeast Asia, observing pronounced allele-frequency differences between population groups. Together, these findings refine models of oceanic settlement and admixture and establish a comprehensive reference that will advance global efforts to ensure broad and equitable representation in human genomics.

q-bio.PE↗

arXiv2025

Efficient Chromosome Parallelization for Precision Medicine Genomic Workflows

Large-scale genomic workflows used in precision medicine can process datasets spanning tens to hundreds of gigabytes per sample, leading to high memory spikes, intensive disk I/O, and task failures due to out-of-memory errors. Simple static resource allocation methods struggle to handle the variability in per-chromosome RAM demands, resulting in poor resource utilization and long runtimes. In this work, we propose multiple mechanisms for adaptive, RAM-efficient parallelization of chromosome-level bioinformatics workflows. First, we develop a symbolic regression model that estimates per-chromosome memory consumption for a given task and introduces an interpolating bias to conservatively minimize over-allocation. Second, we present a dynamic scheduler that adaptively predicts RAM usage with a polynomial regression model, treating task packing as a Knapsack problem to optimally batch jobs based on predicted memory requirements. Additionally, we present a static scheduler that optimizes chromosome processing order to minimize peak memory while preserving throughput. Our proposed methods, evaluated on simulations and real-world genomic pipelines, provide new mechanisms to reduce memory overruns and balance load across threads. We thereby achieve faster end-to-end execution, showcasing the potential to optimize large-scale genomic workflows.

cs.DC↗

arXiv2022

Ancestry-specific analyses of genome-wide data confirm the settlement sequence of Polynesia

By demonstrating the role that historical population replacements and waves of admixture have played around the world, the genetics work of Reich and colleagues has provided a paradigm for understanding human history [Reich et al. 2009; Reich et al. 2012; Patterson et al. 2012]. Although we show in Ioannidis et al. [2021] that the peopling of Polynesia was a range expansion, and not, as suggested by Huang et al. [2022], yet another example of waves of admixture and large-scale gene flow between populations, we believe that our result in this recently settled oceanic expanse is the exception that proves the rule.

q-bio.PE↗

arXiv2019

LitGen: Genetic Literature Recommendation Guided by Human Explanations

As genetic sequencing costs decrease, the lack of clinical interpretation of variants has become the bottleneck in using genetics data. A major rate limiting step in clinical interpretation is the manual curation of evidence in the genetic literature by highly trained biocurators. What makes curation particularly time-consuming is that the curator needs to identify papers that study variant pathogenicity using different types of approaches and evidences---e.g. biochemical assays or case control analysis. In collaboration with the Clinical Genomic Resource (ClinGen)---the flagship NIH program for clinical curation---we propose the first machine learning system, LitGen, that can retrieve papers for a particular variant and filter them by specific evidence types used by curators to assess for pathogenicity. LitGen uses semi-supervised deep learning to predict the type of evidence provided by each paper. It is trained on papers annotated by ClinGen curators and systematically evaluated on new test data collected by ClinGen. LitGen further leverages rich human explanations and unlabeled data to gain 7.9%-12.6% relative performance improvement over models learned only on the annotated papers. It is a useful framework to improve clinical variant curation.

cs.CL↗

arXiv2018

DeepTag: inferring all-cause diagnoses from clinical notes in under-resourced medical domain

Large scale veterinary clinical records can become a powerful resource for patient care and research. However, clinicians lack the time and resource to annotate patient records with standard medical diagnostic codes and most veterinary visits are captured in free text notes. The lack of standard coding makes it challenging to use the clinical data to improve patient care. It is also a major impediment to cross-species translational research, which relies on the ability to accurately identify patient cohorts with specific diagnostic criteria in humans and animals. In order to reduce the coding burden for veterinary clinical practice and aid translational research, we have developed a deep learning algorithm, DeepTag, which automatically infers diagnostic codes from veterinary free text notes. DeepTag is trained on a newly curated dataset of 112,558 veterinary notes manually annotated by experts. DeepTag extends multi-task LSTM with an improved hierarchical objective that captures the semantic structures between diseases. To foster human-machine collaboration, DeepTag also learns to abstain in examples when it is uncertain and defers them to human experts, resulting in improved performance. DeepTag accurately infers disease codes from free text even in challenging cross-hospital settings where the text comes from different clinical settings than the ones used for training. It enables automated disease annotation across a broad range of clinical diagnoses with minimal pre-processing. The technical framework in this work can be applied in other medical domains that currently lack medical coding resources.

cs.CL↗

arXiv2018

Network Enhancement: a general method to denoise weighted biological networks

Networks are ubiquitous in biology where they encode connectivity patterns at all scales of organization, from molecular to the biome. However, biological networks are noisy due to the limitations of measurement technology and inherent natural variation, which can hamper discovery of network patterns and dynamics. We propose Network Enhancement (NE), a method for improving the signal-to-noise ratio of undirected, weighted networks. NE uses a doubly stochastic matrix operator that induces sparsity and provides a closed-form solution that increases spectral eigengap of the input network. As a result, NE removes weak edges, enhances real connections, and leads to better downstream performance. Experiments show that NE improves gene function prediction by denoising tissue-specific interaction networks, alleviates interpretation of noisy Hi-C contact maps from the human genome, and boosts fine-grained identification accuracy of species. Our results indicate that NE is widely applicable for denoising biological networks.

q-bio.MN↗

arXiv2013

Reconstructing Native American Migrations from Whole-genome and Whole-exome Data

There is great scientific and popular interest in understanding the genetic history of populations in the Americas. We wish to understand when different regions of the continent were inhabited, where settlers came from, and how current inhabitants relate genetically to earlier populations. Recent studies unraveled parts of the genetic history of the continent using genotyping arrays and uniparental markers. The 1000 Genomes Project provides a unique opportunity for improving our understanding of population genetic history by providing over a hundred sequenced low coverage genomes and exomes from Colombian (CLM), Mexican-American (MXL), and Puerto Rican (PUR) populations. Here, we explore the genomic contributions of African, European, and Native American ancestry to these populations. Estimated Native American ancestry is 48% in MXL, 25% in CLM, and 13% in PUR. Native American ancestry in PUR is most closely related to populations surrounding the Orinoco River basin, confirming the Southern America ancestry of the Taíno people of the Caribbean. We present new methods to estimate the allele frequencies in the Native American fraction of the populations, and model their distribution using a demographic model for three ancestral Native American populations. These ancestral populations likely split in close succession: the most likely scenario, based on a peopling of the Americas 16 thousand years ago (kya), supports that the MXL Ancestors split 12.2kya, with a subsequent split of the ancestors to CLM and PUR 11.7kya. The model also features effective populations of 62,000 in Mexico, 8,700 in Colombia, and 1,900 in Puerto Rico. Modeling Identity-by-descent and ancestry tract length, we show that post-contact populations differ markedly in their effective sizes and migration patterns, with Puerto Rico showing the smallest effective size and the earlier migration from Europe.

q-bio.PE↗

arXiv2013

Genetic recombination is targeted towards gene promoter regions in dogs

The identification of the H3K4 trimethylase, PRDM9, as the gene responsible for recombination hotspot localization has provided considerable insight into the mechanisms by which recombination is initiated in mammals. However, uniquely amongst mammals, canids appear to lack a functional version of PRDM9 and may therefore provide a model for understanding recombination that occurs in the absence of PRDM9, and thus how PRDM9 functions to shape the recombination landscape. We have constructed a fine-scale genetic map from patterns of linkage disequilibrium assessed using high-throughput sequence data from 51 free-ranging dogs, Canis lupus familiaris. While broad-scale properties of recombination appear similar to other mammalian species, our fine-scale estimates indicate that canine highly elevated recombination rates are observed in the vicinity of CpG rich regions including gene promoter regions, but show little association with H3K4 trimethylation marks identified in spermatocytes. By comparison to genomic data from the Andean fox, Lycalopex culpaeus, we show that biased gene conversion is a plausible mechanism by which the high CpG content of the dog genome could have occurred.

q-bio.PE↗

arXiv2013

Genome Sequencing Highlights Genes Under Selection and the Dynamic Early History of Dogs

To identify genetic changes underlying dog domestication and reconstruct their early evolutionary history, we analyzed novel high-quality genome sequences of three gray wolves, one from each of three putative centers of dog domestication, two ancient dog lineages (Basenji and Dingo) and a golden jackal as an outgroup. We find dogs and wolves diverged through a dynamic process involving population bottlenecks in both lineages and post-divergence gene flow, which confounds previous inferences of dog origins. In dogs, the domestication bottleneck was severe involving a 17 to 49-fold reduction in population size, a much stronger bottleneck than estimated previously from less intensive sequencing efforts. A sharp bottleneck in wolves occurred soon after their divergence from dogs, implying that the pool of diversity from which dogs arose was far larger than represented by modern wolf populations. Conditional on mutation rate, we narrow the plausible range for the date of initial dog domestication to an interval from 11 to 16 thousand years ago. This period predates the rise of agriculture, implying that the earliest dogs arose alongside hunter-gathers rather than agriculturists. Regarding the geographic origin of dogs, we find that surprisingly, none of the extant wolf lineages from putative domestication centers are more closely related to dogs, and the sampled wolves instead form a sister monophyletic clade. This result, in combination with our finding of dog-wolf admixture during the process of domestication, suggests a re-evaluation of past hypotheses of dog origin is necessary. Finally, we also detect signatures of selection, including evidence for selection on genes implicated in morphology, metabolism, and neural development. Uniquely, we find support for selective sweeps at regulatory sites suggesting gene regulatory changes played a critical role in dog domestication.

q-bio.GN↗

arXiv2013

Reconstructing the Population Genetic History of the Caribbean

The Caribbean basin is home to some of the most complex interactions in recent history among previously diverged human populations. Here, by making use of genome-wide SNP array data, we characterize ancestral components of Caribbean populations on a sub-continental level and unveil fine-scale patterns of population structure distinguishing insular from mainland Caribbean populations as well as from other Hispanic/Latino groups. We provide genetic evidence for an inland South American origin of the Native American component in island populations and for extensive pre-Columbian gene flow across the Caribbean basin. The Caribbean-derived European component shows significant differentiation from parental Iberian populations, presumably as a result of founder effects during the colonization of the New World. Based on demographic models, we reconstruct the complex population history of the Caribbean since the onset of continental admixture. We find that insular populations are best modeled as mixtures absorbing two pulses of African migrants, coinciding with early and maximum activity stages of the transatlantic slave trade. These two pulses appear to have originated in different regions within West Africa, imprinting two distinguishable signatures in present day Afro-Caribbean genomes and shedding light on the genetic impact of the dynamics occurring during the slave trade in the Caribbean.

q-bio.PE↗

arXiv2009

Inferring the joint demographic history of multiple populations from multidimensional SNP frequency data

Demographic models built from genetic data play important roles in illuminating prehistorical events and serving as null models in genome scans for selection. We introduce an inference method based on the joint frequency spectrum of genetic variants within and between populations. For candidate models we numerically compute the expected spectrum using a diffusion approximation to the one-locus two-allele Wright-Fisher process, involving up to three simultaneous populations. Our approach is a composite likelihood scheme, since linkage between neutral loci alters the variance but not the expectation of the frequency spectrum. We thus use bootstraps incorporating linkage to estimate uncertainties for parameters and significance values for hypothesis tests. Our method can also incorporate selection on single sites, predicting the joint distribution of selected alleles among populations experiencing a bevy of evolutionary forces, including expansions, contractions, migrations, and admixture. As applications, we model human expansion out of Africa and the settlement of the New World, using 5 Mb of noncoding DNA resequenced in 68 individuals from 4 populations (YRI, CHB, CEU, and MXL) by the Environmental Genome Project. We also combine our demographic model with a previously estimated distribution of selective effects among newly arising amino acid mutations to accurately predict the frequency spectrum of nonsynonymous variants across three continental populations (YRI, CHB, CEU).

q-bio.PE↗