BrentLab/mahendrawada_2025

Mahendrawada 2025

This data is taken from the Supplement of

Mahendrawada, L., Warfield, L., Donczew, R. et al. Low overlap of transcription factor DNA binding and regulatory targets. Nature 642, 796–804 (2025). https://doi.org/10.1038/s41586-025-08916-0

Please note that all column descriptions, except for those containing "locus_tag" and "symbol", are copy/pasted directly from the supplement or methods section of this paper.

Dataset Details

This repo provides three datasets:

• genomic_features: These are the genomic features provided in Supplemental Table 2
• chec_seq: binding location data provided in Supplemental Table 3. the peak_score is defined to be the ChEC signal around peak center' (sum of ChEC signal from -150 to +150 bp from peak summit) normalized to Drosophila spike-in control. How these scores are assigned to features isn't well described in either the supplement or Methods section, but it is reasonable to believe that it is done with HOMER.
• rna_seq: Nascent RNA-seq differential expression data following transcription factor depletion using 4TU metabolic labeling filtered for significantly affected genes (DESeq2, padj <0.1, FC >= 1.3)

Data Structure

This dataset provides the following three parquet files

ChEC-seq

Field	Description
regulator_locus_tag	Systematic gene name (ORF identifier) of the transcription factor
regulator_symbol	Standard gene symbol of the transcription factor
target_locus_tag	Systematic gene name (ORF identifier) of the target gene
target_symbol	Standard gene symbol of the target gene
peak_score	ChEC signal around peak center (sum of ChEC signal from -150 to +150 bp from peak summit) normalized to Drosophila spike-in control

RNA-seq

Field	Description
regulator_locus_tag	Systematic gene name (ORF identifier) of the depleted transcription factor
regulator_symbol	Standard gene symbol of the depleted transcription factor
target_locus_tag	Systematic gene name (ORF identifier) of the differentially expressed target gene
target_symbol	Standard gene symbol of the differentially expressed target gene
log2fc	Log2 fold change (IAA/DMSO) for significantly affected genes (DESeq2, padj <0.1, FC >= 1.3)

Features

Field	Description
gene_id	Systematic gene name (ORF identifier) from SGD (https://yeastgenome.org/)
SGD_id	Unique identifier for each gene from SGD (https://yeastgenome.org/)
gene_name	Common name of each gene
chr	Chromosome number corresponding to gene
strand	Strandedness of the gene (+ or -)
start	Start position of the ORF
end	End position of the ORF
TSS	Transcription start site based on Park et al., 2014 (doi:10.1093/nar/gkt1366)
TATA_category	TATA box classification from Donczew et al., 2020 using consensus TATAWAW (doi:10.7554/eLife.50109)
expression	Average signal normalized to gene length from Donczew et al., 2020 (doi:10.7554/eLife.50109)
+1 nucleosome	Position of +1 nucleosome from Chereji et al., 2018 (doi:10.1186/S13059-018-1398-0)
-1 nucleosome	Position of -1 nucleosome from Chereji et al., 2018 (doi:10.1186/S13059-018-1398-0)
NDR Center	Center of nucleosome depleted region from Chereji et al., 2018 (doi:10.1186/S13059-018-1398-0)
NDR Width	Width of nucleosome depletion region from Chereji et al., 2018 (doi:10.1186/S13059-018-1398-0)
tail-dependence	Tail classification based on Mediator tail dependence from Warfield L, Donczew R et al., 2022 (doi:10.1016/j.molcel.2022.09.016)
coactivator	Coactivator classification based on TFIID and/or SAGA dependence from Donczew et al., 2020 (doi:10.7554/eLife.50109)
LCID_center	Genes near boundaries of chromosomal interacting domains from Swygert et al., 2020 (doi:10.1016/j.molcel.2018.11.020)
Rossi_classes	Promoter classes from Rossi et al., 2021 (doi:10.1038/s41586-021-03314-8)
RP_category	Ribosomal protein (RP) and ribosomal biogenesis (RiBi) gene classification from Zencir et al., 2020 (doi:10.1093/NAR/GKAA852)
binding_cluster	Clusters from unsupervised K-means clustering using binary binding data of 178 transcription factors
list_of_TFS_bound	List of transcription factors bound to gene promoter (-400 to +200 bp from TSS; Homer peak calling)
number_of_bound_tfs	Number of transcription factors bound to each promoter
locus_tag	Systematic gene identifier from yeast_genome_resources dataset
symbol	Standard gene symbol from yeast_genome_resources dataset

Usage

There are three parquet files in this repo. This is a way of getting that information programmatically

from huggingface_hub import ModelCard
from pprint import pprint

card = ModelCard.load("BrentLab/hughes_2006", 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)

With the result

{'chec_seq': 'chec_mahendrawada_2025.parquet',
 'genomic_features': 'features_mahendrawada_2025.parquet',
 'rna_seq': 'rnaseq_mahendrawada_2025.parquet'}

I recommend using huggingface_hub.snapshot_download to pull the repository. After that, use your favorite method of interacting with parquet files (eg duckDB, but you could use dplyr in R or pandas, too).

from huggingface_hub import snapshot_download
import duckdb
import os

repo_id = "BrentLab/mahendrawada_2025"

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

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

# Construct path to the checseq parquet file
parquet_path = os.path.join(repo_path, "chec_mahendrawada_2025.parquet")
print(f"Parquet file at: {parquet_path}")

# Connect to DuckDB and query the parquet file
conn = duckdb.connect()

query = """
SELECT * 
FROM read_parquet(?)
WHERE regulator_symbol = 'CST6'
"""

result = conn.execute(query, [parquet_path]).fetchall()
print(f"Found {len(result)} rows for CST6")