Skip to main content

Documentation Index

Fetch the complete documentation index at: https://mintlify.com/BDB-Genomics/AlphaGenomeR/llms.txt

Use this file to discover all available pages before exploring further.

AlphaGenomeR is a Bioconductor-ready R package that gives R researchers direct access to Google DeepMind’s AlphaGenome API without leaving their native R environment. It handles all Python interop internally via reticulate, so you work entirely in R while the package manages gRPC communication with the AlphaGenome service and converts results into standard matrices and data frames.

What is AlphaGenomeR?

AlphaGenome is a multimodal transformer model developed by Google DeepMind that predicts functional genomic features directly from DNA sequences at single-base resolution. Given a 1 MB genomic interval, it simultaneously predicts RNA-seq signal, chromatin accessibility, transcription factor binding, splicing patterns, and 3D chromatin contact maps — filtered by tissue or cell type. AlphaGenomeR bridges the official AlphaGenome Python SDK using the reticulate package. You call R functions, AlphaGenomeR translates your inputs into Python objects, executes the gRPC prediction request, and returns the results as R-native matrix and data.frame objects ready for downstream analysis with standard Bioconductor tools.

Key features

Multimodal predictions

Query 11+ genomic modalities in a single API call — RNA-seq, ATAC-seq, CAGE, ChIP-seq (TF and histone), DNase, splice sites, splice junctions, PRO-cap, and 3D contact maps.

Bioconductor integration

Results are returned as R-native matrix and data.frame objects compatible with standard Bioconductor analysis workflows.

Tissue specificity

Filter predictions by tissue or cell type using UBERON and CL ontology terms (e.g., "UBERON:0002048" for lung).

High resolution

Predictions span 1 MB genomic intervals at single-base pair resolution for precise regulatory analysis.

Prerequisites

AlphaGenomeR requires the following before installation:
  • Python >= 3.10 — used by reticulate to load the AlphaGenome SDK
  • alphagenome Python package — the official SDK that handles gRPC communication with the API
  • AlphaGenome API key — a free, non-commercial key available from the AlphaGenome Science Page

How it works

When you call alphagenome_query(), the package uses reticulate to import the alphagenome.models.dna_client and alphagenome.data.genome Python modules. It parses your genomic region string, maps R character vectors to Python enum values, initializes a gRPC client with your API key, and calls client$predict_interval(). The raw Python response object is then converted to an R list via reticulate::py_to_r(). Extractor functions like alphagenome_get_rna_seq() and alphagenome_get_atac() pull specific modalities out of that list and return a named list with two elements: values (a numeric matrix of positions × tracks) and metadata (a data frame describing each track).

Next steps

Installation

Install AlphaGenomeR from GitHub and set up the Python dependency.

Quickstart

Make your first multimodal genomic prediction in minutes.

Build docs developers (and LLMs) love