snapatac2.tl.init_network_from_annotation#

snapatac2.tl.init_network_from_annotation(regions, anno_file, upstream=250000, downstream=250000, id_type='gene_name', coding_gene_only=True)[source]#

Build a region-to-gene network from gene annotations.

Use this function to connect candidate cis-regulatory elements to genes when the regions fall within an annotation-derived regulatory domain around each transcription start site.

Anti-Patterns#

Do NOT pass regions from a genome build different from anno_file.
Do NOT assume edges are functional regulatory links; they encode genomic proximity only until scores are added.

param regions:: Candidate regulatory regions in chrom:start-end format.
type regions:: list[str]
param anno_file:: GFF/GTF annotation file, or a Genome object containing the annotation.
type anno_file:: Path | Genome
param upstream:: Bases upstream of each transcription start site included in the regulatory domain.
type upstream:: int
param downstream:: Bases downstream of each transcription start site included in the regulatory domain.
type downstream:: int
param id_type:: Annotation identifier stored on gene nodes.
type id_type:: Literal['gene_name', 'gene_id', 'transcript_id']
param coding_gene_only:: If True, retain only protein-coding genes.
type coding_gene_only:: bool
returns:: Directed graph whose region nodes point to nearby gene nodes.
rtype:: PyDiGraph

Examples

>>> import snapatac2 as snap
>>> regions = ["chr1:10000-10500", "chr1:20000-20500"]
>>> network = snap.tl.init_network_from_annotation(regions, snap.genome.hg38)
>>> network.num_nodes() >= 0
True