Example datasets

The combined LAML oncoplot uses load_dataset("tcga_laml_combined_oncoplot"); the p53 sequence comparison loads aligned FASTA text with load_dataset("p53_sequence_comparison", as_format="text"). Their sources and processing are described in the combined oncoplot and p53 comparison gallery pages.

The package ships the tables that the gallery examples use, so you can try the API on real data without downloading anything.

Data retain their own terms, independently of the source packages’ software licenses. See the third-party notices for file-level status, attribution, and terms.

from genome_spy.datasets import available_datasets, load_dataset

available_datasets()
counts = load_dataset("airway_scaledcounts")

load_dataset() returns a pandas DataFrame for tabular files and parsed JSON for JSON files. Pass as_format="text" to get the raw file contents instead.

Dataset

Contents

airway_review

Gallery airway statistics and sample counts for downstream gene review

airway_metadata

Sample table for the airway RNA-seq experiment

airway_scaledcounts

Rounded, length-scaled gene counts for the same eight samples

hapmap_gwas

HapMap coordinates with simulated p-values and effect sizes

brca_maf

Somatic mutation calls for one TCGA breast-tumor sample

pik3ca_mutations

Historical UniProt feature counts and Pfam protein domains

tcga_oncoprint

Plotly Dash Bio alteration fixture

mutation_impact_reference

Project-authored synthetic reference window

tal1_alphagenome_reference

UCSC hg38 reference sequence with interval metadata

pik3ca_tcga_brca_lollipop

Prepared recurrent PIK3CA mutations and protein domains

tcga_laml_maf

Somatic mutation calls for TCGA acute myeloid leukemia

tcga_laml_annotations

Clinical annotations for those leukemia samples

tcga_laml_combined_oncoplot

Prepared mutation, copy-number, clinical, pathway, VAF, and MutSig tables

p53_sequence_comparison

34 p53 protein sequences aligned with MAFFT L-INS-i, compressed FASTA

pyoncoprint_tcga

Alteration matrix for TCGA lung adenocarcinoma samples

tcga_ov_gistic_scores

GISTIC2 copy-number scores for TCGA ovarian tumors

tcga_ov_gistic_lesions

GISTIC2 peak regions for the same cohort

refseq_gene_bodies

Assembly-wide hg19 and hg38 RefSeq gene bodies

rnf7_direct_rna

Prepared RNF7 direct-RNA coverage, reference bases, read-alignment events, exon intervals, and m6Anet site probabilities

Data use and provenance

  • airway_metadata and airway_scaledcounts describe the airway smooth muscle RNA-seq experiment of Himes et al., PLoS One 2014 (GEO GSE52778), independently reprocessed by Stephen Turner for the Bioconnector workshops with kallisto and tximport lengthScaledTPM. The bundled CSVs exactly match those workshop files; the workshop declares CC BY-NC-SA 4.0.

  • hapmap_gwas is the HapMap example table from the manhattanly R package. HapMap supplied the build-36 map coordinates and rs identifiers; UCSC hg18 supplied gene annotations. The association statistics are simulated or derived from simulated p-values. The table contains no individual genotypes.

  • brca_maf, tcga_laml_maf, and tcga_laml_annotations are the example files bundled with maftools and contain TCGA mutation calls or clinical annotations.

  • pik3ca_tcga_brca_lollipop contains the chart-ready named datasets from GenomeSpy’s official TCGA-BRCA PIK3CA lollipop example. Mutation counts come from GDC masked somatic MAFs and protein domains from UniProt P42336; see the gallery example for full provenance.

  • pyoncoprint_tcga is the example alteration table from pyoncoprint as a cBioPortal TCGA LUAD export. See the cBioPortal data policy and the study-specific notice linked in our third-party notices. Its microbiome track comes from a study retracted in 2024 and is not a validated finding.

  • tcga_ov_gistic_scores and tcga_ov_gistic_lesions are the complete scores.gistic and all_lesions.conf_99.txt tables used by the official GenomeSpy example. They are open-access TCGA OV-TP GISTIC2 output produced by the Broad Institute TCGA Genome Data Analysis Center, Firehose run 2016-01-28 (source archive).

  • refseq_gene_bodies is independently prepared from the official hg19 and hg38 UCSC refGene tables. Overlapping transcripts are collapsed by symbol, chromosome, and strand; transcript counts prioritize colliding labels. See UCSC’s conditions for use.

  • rnf7_direct_rna combines a reduced transcript-window extract from the CC BY 4.0 xPore demo archive, RNF7 rows from m6Anet Supplementary Table 6, and three GRCh38 exon sequences from Ensembl. Python preparation pseudonymized molecule labels, calculated coverage, and derived mismatches and CIGAR events for the direct-RNA gallery example.

Results shown with the TCGA-derived tables are in whole or part based upon data generated by the TCGA Research Network.