# Isoform-resolved Mean Ribosome Load (MRL) of human kidney cancer cell transcripts, Sugimoto and Ratcliffe 2022 (mRNABench)

## Facts

- Provider: Yoichiro Sugimoto and Peter J. Ratcliffe (Oxford); packaged by the Morris lab (MSKCC) for mRNABench
- License: CC-BY-4.0 (https://www.nature.com/articles/s41594-022-00819-2)
- Size: 3 files
- Format: Parquet (in sampled files)
- Stored at: Hugging Face, https://huggingface.co/datasets/morrislab/mrl-sugimoto
- Access: public, readable without an account
- Category: Genomics & Bioinformatics
- Source: https://doi.org/10.1038/s41594-022-00819-2
- Page: https://scigantic.com/archive/b8632c0f-26cb-5a66-8235-b147a7e3a173
- Keywords: Genomics & Bioinformatics, benchmark, Parquet, Hugging Face

## About

Mean ribosome load (MRL), the average number of ribosomes bound to an mRNA and a measure of how actively it is translated, measured per transcript isoform in human kidney cancer cells by Sugimoto and Ratcliffe, who combined polysome profiling with long-read isoform-resolved sequencing to study how HIF and mTOR dysregulation reshape translation. The Morris lab repackaged it as a regression task in mRNABench, the benchmark for mRNA language models. mrl-sugimoto.parquet has 11,693 rows, one per transcript isoform, with gene name, chromosome, the full transcript sequence, a binary track marking the first nucleotide of each codon in the coding sequence (CDS), a binary track marking exon 3' ends (splice sites) and the MRL target; about 12 MB. Read with pandas, polars or Hugging Face datasets, or through the mrnabench Python package. Use for predicting translation efficiency from mRNA sequence, evaluating RNA foundation model embeddings and studying 5' UTR and isoform effects on ribosome loading. The source data are published with the CC BY 4.0 open access article; cite Sugimoto and Ratcliffe, Nature Structural and Molecular Biology 29, 871-880 (2022), doi 10.1038/s41594-022-00819-2, and mRNABench.

## Load the data

```python
from datasets import load_dataset
ds = load_dataset("morrislab/mrl-sugimoto")
print(ds)
```

## Data structure

File types in sampled files: Parquet

Example file: `mrl-sugimoto.parquet`

## Open it

- In a hosted notebook with the data mounted: https://scigantic.com/archive/b8632c0f-26cb-5a66-8235-b147a7e3a173
- Full record as JSON (schema card included): https://api.scigantic.com/api/archives/b8632c0f-26cb-5a66-8235-b147a7e3a173
- Over MCP: `get_schema_card` with id `b8632c0f-26cb-5a66-8235-b147a7e3a173` at https://api.scigantic.com/mcp

## Related topics

- [Parquet datasets](https://scigantic.com/topics/parquet)
- [Genomics & Bioinformatics datasets](https://scigantic.com/topics/genomics-and-bioinformatics)
- [Benchmark datasets](https://scigantic.com/topics/benchmark)
