Yeast RNA and ribosome occupancy

ingolia-yeast-rna-ribo contains the processed RNA and ribosome-footprint measurements from Ingolia et al. (2009), GSE13750. It uses the same dataset API as human, mouse and dog.

pip install 'orfdata[data]'
orfdata datasets info ingolia-yeast-rna-ribo
orfdata datasets download ingolia-yeast-rna-ribo

The download is about 24 MB compressed, 210 MB installed. It contains 329,952 observations: counts, reads/kb and RPKM for 13,748 source regions in eight libraries. Four libraries measure RNA and four measure ribosome footprints, in rich medium or after 20 minutes of amino-acid starvation. Source sample titles identify the modality, condition and nominal replicate. In particular, GSM346114 is Footprints-rich-2 despite its fp_rich1 filename.

from orfdata import DatasetRegistry, ScopedTarget

dataset = DatasetRegistry().open("ingolia-yeast-rna-ribo")
reference, = dataset.manifest.references
target = ScopedTarget(reference, "region", "HRA1", "Ingolia2009:feature")
for measurement in dataset.measurements(target=target, unit="RPKM"):
    print(measurement.sample_id, measurement.quantity, measurement.value)

Queries work offline after download. quantity="rna_abundance" or quantity="ribosome_occupancy" selects one modality. unit="counts" returns the original aligned read counts. The other two native series have origin="derived" and link to their count observation. Source formulas, feature lengths and total CDS-aligned reads remain attached; the reported rounded values are preserved. Zeros remain measured zeros. GSM346118 has no zero-count rows; ORFdata does not add or remove a pseudocount.

What these regions establish

The source includes 3,523 5′ UTRs, 4,592 3′ UTRs, 225 intron windows and 5,408 other feature labels, including coding, ncRNA and mitochondrial labels. Ribosome occupancy over a UTR does not identify a particular uORF. These targets therefore retain kind="region"; they do not expand the experimental ORF catalogue or establish stable protein abundance, copies/cell or translation flux.

The yeast reference has taxon 4932 and the source annotation date, SGD 2008-06-22. Exact genomic windows and assembly remain unresolved, including the UTR definitions attributed to Nagalakshmi et al. (2008). The biological strain BY4741 is sample metadata. Current PyEnsembl coordinates cannot be substituted for these historical windows without a validated mapping.

The paper reports RNA and footprints from the same samples, but the exact GSM-to-GSM aliquot crosswalk is unresolved. The release retains eight separate libraries and creates no assay pairs or translation-efficiency ratios. It also does not pair these cultures with the separate CoPY protein study. Issue #11 tracks these source gaps.

To rebuild, install orfdata[data,build-data] and run orfdata datasets build ingolia-yeast-rna-ribo, or DatasetRegistry().build("ingolia-yeast-rna-ribo"). All 17 source inputs have pinned URLs, sizes and checksums. Verified local inputs can be supplied through --source-path ASSET=PATH or source_paths={...}. Query a local build using DatasetRegistry().open_build(...) or orfdata datasets query ... --built.