+Every person's genome is a mosaic inherited from ancestors of different +populations, and "local ancestry" describes, region by region along a chromosome, +where a stretch of DNA came from. Point cloud local ancestry inference (pcLAI) +does this without picking one population label per region: it places each region +at a point in a continuous coordinate space, so ancestry that sits between two +reference populations stays visible as an intermediate position instead of being +rounded to the nearest label. +
++This track shows pcLAI for Human Pangenome Reference Consortium (HPRC) Release 2, +projected from each assembly onto GRCh38: both haplotypes of 231 samples, plus +CHM13, so 463 subtracks in all. Each haplotype is divided into windows of 1000 +SNPs, which works out to roughly a hundred thousand bases, about 26,000 windows +per haplotype and 11.9 million in total. Only the autosomes are covered. +
+ ++Each item is one genomic window. Items carry no visible label. Holding the mouse +over an item shows the haplotype, the window identifier, the position of the +window in the principal-component space that pcLAI uses to describe ancestry, the +ancestry the window is assigned to if the continuous coordinate is discretized, +and a confidence score; the same values are on the details page, together with a +scatterplot +showing where the window sits relative to the reference haplotypes that define +the space. +
++The item color is derived from the principal-component position, so windows of +similar inferred ancestry get similar colors and a run of shared ancestry appears +as a block of consistent color. Because the color is continuous rather than a set +of population labels, it is read by comparing regions with each other rather than +against a fixed legend. Neighboring windows inside one block differ by only one or +two color steps, and those small differences are not meaningful: the feature to +read is the block, not the window. +
++Two consequences are worth knowing before concluding that a display looks wrong. +First, a haplotype with a single ancestry throughout gives one uniform color +across every chromosome, which is the correct result for that sample rather than a +rendering problem; 48 of the 463 haplotypes are like this. Second, where a +haplotype does carry several ancestries the blocks are tens of megabases long, so +a view of only a few megabases usually falls inside one block and also looks +uniform. Zooming out to a whole chromosome is what makes the block structure +visible. +
++The track opens with seven haplotypes turned on. Five of them were picked because +their ancestry is visibly mixed: NA19682 (both haplotypes), NA19776, HG01496 and +HG01150 each carry a second ancestry over more than a quarter of the genome. The +other two, HG002 and CHM13, are there as familiar reference points and are close +to uniform. The remaining haplotypes can be turned on from the sample-by-haplotype +matrix on the track configuration page. +
++Dense mode is the mode to use. It draws one row per haplotype, and because the +items carry no label nothing is lost by collapsing them: a window is still clicked +or moused over for its values. Pack mode adds nothing here and costs a lot of +vertical space, since the windows abut and pack puts each one on a different row +from its neighbour. +
+ ++Conventional local ancestry inference gives every segment of a genome one of a +fixed set of population labels. pcLAI instead places each segment at a point in a +continuous coordinate space, so a genome becomes a cloud of points, one per +window. The coordinate space can be any continuous description of ancestry; here +it is the first two principal components of a reference PCA embedding built from +3122 1000 Genomes haplotypes of 21 populations of known origin. Each haplotype was +split into windows of 1000 SNPs and a coordinate predicted for each, together with +a confidence score; windows of very low confidence were dropped before the BED +files were published, which together with the projection onto GRCh38 is why the +windows do not tile a chromosome without gaps (for NA19682 haplotype 1, 2.73 of +the 2.88 Gb of autosome is covered, with about 65 Mb of internal gaps). The files +also +carry a discretized version of each prediction, written as the PCA centroid of the +ancestry cluster the window falls in, and across all 463 haplotypes that takes +only four values. See the reference below for the method, and the +pcLAI repository +for the code and the reference panel table. +
++The annotation files were obtained from the HPRC Release 2 data collection on the +public s3://human-pangenomics bucket, indexed at +the hprc_intermediate_assembly data tables; +this track uses the GRCh38-coordinate index, +pclai_v1.1_grch38_coord_local_hprc_r2.index.csv. Each per-haplotype BED +file was converted to a UCSC bigBed file, with the haplotype, the window +identifier and the window's own principal-component coordinate split out of the +source item name into their own fields, and the source's centroid column named. All 11,936,603 windows in the source files +are carried through to the bigBeds; none were dropped. In 54,811 of them (0.46%) +the source thickStart is one base before chromStart and so +outside the item, which is invalid BED and also departs from the format the pcLAI +authors document; since thickStart and +thickEnd carry no information for this annotation they were set to the +item bounds rather than dropping those windows. The steps are described in the +makeDoc +and the build scripts are in the +kent source tree. +
+ ++The data can be explored interactively in table format with the +Table Browser or the +Data Integrator and exported from there to +spreadsheet or tab-sep tables. From scripts, the data can be accessed through our +API, +track=hprcPclai. +
++For automated download and analysis, the annotation is stored in one bigBed file +per haplotype, which can be downloaded from +our download server. +The files are called SAMPLE.HAPLOTYPE.bb, for example +NA19682.1.bb. Individual regions or the whole genome annotation can be +obtained using our tool bigBedToBed, which can be compiled from the +source code or downloaded as a precompiled binary for your system. Instructions +for downloading source code and binaries can be found +here. +The tool can also be used to obtain features within a given range, e.g. +bigBedToBed http://hgdownload.soe.ucsc.edu/gbdb/hg38/hprcPclai/NA19682.1.bb -chrom=chr21 -start=0 -end=100000000 stdout +
++The original annotation source data can be downloaded from the HPRC S3 bucket +linked above. +
+ ++Annotations were generated by the Human Pangenome Reference Consortium. Thanks to +the HPRC production team and to the pcLAI authors for making these data available. +
+ ++Geleta M, Mas Montserrat D, Ioannidis NM, Ioannidis AG. + +Point cloud local ancestry inference (PCLAI): continuous coordinate-based ancestry along the +genome. +bioRxiv. 2026 Mar 25. +doi: 10.64898/2026.03.23.713813 +