Accurate Inference of Local Phased Ancestry of Modern Admixed Populations

Abstract

Population stratification is a growing concern in genetic-association studies. Averaged ancestry at the genome level (global ancestry) is insufficient for detecting the population substructures and correcting population stratifications in association studies. Local and phase stratification are needed for human genetic studies, but current technologies cannot be applied on the entire genome data due to various technical caveats. Here we developed a novel approach (aMAP, ancestry of Modern Admixed Populations) for inferring local phased ancestry. It took about 3 seconds on a desktop computer to finish a local ancestry analysis for each human genome with 1.4-million SNPs. This method also exhibits the scalability to larger datasets with respect to the number of SNPs, the number of samples, and the size of reference panels. It can detect the lack of the proxy of reference panels. The accuracy was 99.4%. The aMAP software has a capacity for analyzing 6-way admixed individuals. As the biomedical community continues to expand its efforts to increase the representation of diverse populations, and as the number of large whole-genome sequence datasets continues to grow rapidly, there is an increasing demand on rapid and accurate local ancestry analysis in genetics, pharmacogenomics, population genetics, and clinical diagnosis.

Efficient Genome-Wide Sequencing and Low-Coverage Pedigree Analysis from Noninvasively Collected Samples

“…However, for the many types of analyses that use genome-scale data [ e.g. , local ancestry analysis, genome-wide scans for selection, and reconstruction of population demographic history ( Sabeti et al 2002 ; Huang et al 2007 ; Li et al 2007 ; Sankararaman and Sridhar 2008 ; Price et al 2009 ; Durand et al 2011 ; Li and Durbin 2011 ; Yang et al 2011 ; Ma et al 2014 )], our approach will be more useful, especially as the costs of high-throughput sequencing continue to fall.…”

“…Thus, while existing methods are sufficient for basic pedigree construction and estimating some population genetic parameters (although usually with substantial uncertainty), they are severely underpowered for many other types of analyses ( Sabeti et al 2002 ; Price et al 2009 ; Li and Durbin 2011 ), including any that require local ( i.e. , gene- or region-specific) information instead of genome-wide averages ( Huang et al 2007 ; Li et al 2007 ; Sankararaman and Sridhar 2008 ; Yang et al 2011 ; Ma et al 2014 ). Further, because noninvasively collected genotype data are most often based on microsatellites, they cannot take advantage of new tools designed specifically for single-nucleotide variants ( Purcell et al 2007 ; Visscher 2009 ; Durand et al 2011 ).…”

The Mirror RNA Expression Pattern in Human Tissues

“…Then we used a threshold of =<50 as a cutoff to filter away the noise expressions. The remaining expression probes were organized by chromosome positions and orientations [24] .…”

Chromosome-Range Whole-Genome High-Throughput Experimental Haplotyping by Single-Chromosome Microdissection

“…3). Besides the features on high-accuracy, high-throughput, low-cost, and applicability to all types of genetic markers [13, 17, 30–32], a unique feature of this technology is the extremely long length (or full chromosome-length) of the recovered haplotype.…”