Package: GenomeScaleEmbeddings 0.0.0.9000

GenomeScaleEmbeddings: Exploring Genome-Scale Embeddings in 'Parquet' Files

Provides tools to explore genome-scale variant embeddings stored in 'Parquet' files, including data access, matrix conversion, principal component analysis, summaries, and visualization. The datasets accompany Niu et al. (2025), "Incorporating Large Language Model (LLM) Embeddings for Variation Across the Human Genome" <https://arxiv.org/abs/2509.20702>.

Authors:Sounkou Mahamane Toure [aut, cre]

# Install 'GenomeScaleEmbeddings' in R:

install.packages('GenomeScaleEmbeddings', repos = c('https://sounkou-bioinfo.r-universe.dev', 'https://cloud.r-project.org'))

Bug tracker:https://github.com/sounkou-bioinfo/genomescaleembeddings/issues

On CRAN:

2.18 score 2 scripts 13 exports 35 dependencies

Last updated from:c5b01269e2. Checks:9 OK. Indexed: yes.

Target	Result	Time
linux-devel	OK	135
source / vignettes	OK	235
linux-release	OK	138
macos-release	OK	104
macos-oldrel	OK	64
windows-devel	OK	77
windows-release	OK	87
windows-oldrel	OK	72
wasm-release	OK	165

Exports:attachHoubaBigMatrix CopyParquetToDuckDB correlatePCWithPosition DatasetParquetUrlList embeddingSummary getPcaScores houbaPCA infoSummary IterateEmbeddingsMatrixBatches OpenRemoteParquetView plotPcaDims plotPCSpatialCorrelation writeEmbeddingsHoubaFromDuckDB

Dependencies:BH bigmemory bigmemory.sri bigPCAcpp cli cpp11 DBI dplyr duckdb farver generics ggplot2 glue gtable houba isoband jsonlite labeling lifecycle magrittr pillar pkgconfig R6 RColorBrewer Rcpp rlang S7 scales tibble tidyselect utf8 uuid vctrs viridisLite withr

Citation

Development and contributors

Readme and manuals

Help Manual

Help page	Topics
Attach a houba file and return the bigmemory::big.matrix	attachHoubaBigMatrix
Copy remote parquet files into a local DuckDB database file	CopyParquetToDuckDB
Compute correlation between PC scores and genomic position, per chromosome	correlatePCWithPosition
List of the huggingface datasets for the paper "Incorporating LLM Embeddings for Variation Across the Human Genome"	DatasetParquetUrlList
Quick summary for houba mmatrix	embeddingSummary
Get PCA scores from houbaPCA result	getPcaScores
PCA using bigPCAcpp on houba mmatrix or bigmemory::big.matrix	houbaPCA
Quick summary for houba info mmatrix	infoSummary
Iterate over embeddings as matrix batches from a local DuckDB file	IterateEmbeddingsMatrixBatches
Open remote parquet files as a DuckDB VIEW and return as tibble (minimal, http(s) only)	OpenRemoteParquetView
Plot PCA dimensions using ggplot2, colored by annotation	plotPcaDims
Plot spatial correlation between PC scores and genomic position, faceted by chromosome	plotPCSpatialCorrelation
Print a Houba embeddings object	print.HoubaEmbeddings
Write embeddings to houba mmatrix and return info as data.frame from a local DuckDB file	writeEmbeddingsHoubaFromDuckDB