BrentLab/rossi_2021

Rossi 2021

This data is gathered from yeastepigenome.org. This work was published in

Rossi MJ, Kuntala PK, Lai WKM, Yamada N, Badjatia N, Mittal C, Kuzu G, Bocklund K, Farrell NP, Blanda TR, Mairose JD, Basting AV, Mistretta KS, Rocco DJ, Perkinson ES, Kellogg GD, Mahony S, Pugh BF. A high-resolution protein architecture of the budding yeast genome. Nature. 2021 Apr;592(7853):309-314. doi: 10.1038/s41586-021-03314-8. Epub 2021 Mar 10. PMID: 33692541; PMCID: PMC8035251.

This repo provides 4 datasets:

• rossi_2021_metadata: Metadata describing the tagged regulator in each experiment.
• genome_map: ChIP-exo 5' tag coverage data partitioned by sample accession.
• reprocess_annotatedfeatures: This data was reprocessed from the fastq files on GEO. See scripts/reprocessing_details.txt for more information.
• yeastepigenome_annotatedfeatures: ChIP-exo regulator-target binding features with peak statistics.
• reprocess_annotatedfeatures_tagcounts: Reprocessed using a similar method to the calling cards quantification

Usage

The python package tfbpapi provides an interface to this data which eases examining the datasets, field definitions and other operations. You may also download the parquet datasets directly from hugging face by clicking on "Files and Versions", or by using the huggingface_cli and duckdb directly. In both cases, this provides a method of retrieving dataset and field definitions.

tfbpapi

After installing tfbpapi, you can adapt this tutorial in order to explore the contents of this repository.

huggingface_cli/duckdb

You can retrieves and displays the file paths for each configuration of the "BrentLab/rossi_2021" dataset from Hugging Face Hub.

from huggingface_hub import ModelCard
from pprint import pprint

card = ModelCard.load("BrentLab/rossi_2021", repo_type="dataset")

# cast to dict
card_dict = card.data.to_dict()

# Get partition information
dataset_paths_dict = {d.get("config_name"): d.get("data_files")[0].get("path") for d in card_dict.get("configs")}

pprint(dataset_paths_dict)

The entire repository is large. It may be preferable to only retrieve specific files or partitions. You can use the metadata files to choose which files to pull.

from huggingface_hub import snapshot_download
import duckdb
import os
# Download only the metadata first
repo_path = snapshot_download(
    repo_id="BrentLab/rossi_2021",
    repo_type="dataset",
    allow_patterns="rossi_2021_metadata.parquet"
)

dataset_path = os.path.join(repo_path, "rossi_2021_metadata.parquet")
conn = duckdb.connect()
meta_res = conn.execute("SELECT * FROM read_parquet(?) LIMIT 10", [dataset_path]).df()
print(meta_res)

We might choose to take a look at the file with accession SRR11466106:

# Download only a specific sample's genome coverage data
repo_path = snapshot_download(
    repo_id="BrentLab/rossi_2021",
    repo_type="dataset",
    allow_patterns="genome_map/accession=SRR11466106/*.parquet"
)

# Query the specific partition
dataset_path = os.path.join(repo_path, "genome_map")
result = conn.execute("SELECT * FROM read_parquet(?) LIMIT 10", 
                     [f"{dataset_path}/**/*.parquet"]).df()
print(result)

If you wish to pull the entire repo, due to its size you may need to use an authentication token. If you do not have one, try omitting the token related code below and see if it works. Else, create a token and provide it like so:

repo_id = "BrentLab/rossi_2021"

hf_token = os.getenv("HF_TOKEN")

# Download entire repo to local directory
repo_path = snapshot_download(
    repo_id=repo_id,
    repo_type="dataset",
    token=hf_token
)

print(f"\n✓ Repository downloaded to: {repo_path}")

# Construct path to the rossi_annotated_features parquet file
parquet_path = os.path.join(repo_path, "yeastepigenome_annotatedfeatures.parquet")
print(f"✓ Parquet file at: {parquet_path}")