Risk Transcription Factors for Breast and Prostate Cancer with Prioritized Variants

Qing Li's dataset lists risk transcription factors for breast and prostate cancer, derived from a transfer learning framework applied to the Enformer model. The dataset includes tissue-specific cis-regulatory activity scores for millions of single-nucleotide variants prioritized from GWAS datasets. It was last updated on 2026-05-06 and is shared under a CC-BY-4.0 license.

Use Cases

• Prioritizing high-impact regulatory variants for breast and prostate cancer based on tissue-specific cis-regulatory activity scores.
• Identifying cancer-relevant genes for transcriptome-wide association studies using transfer learning-prioritized SNVs.
• Benchmarking tissue-specific model performance against base models for enrichment in enhancers and pathogenic variants.
• Investigating functional essentiality and therapeutic tractability of genes linked to cancer GWAS risk variants.

Strengths

• Derived from a transfer learning framework applied to Enformer, a model trained on 5,313 multi-omics tracks.
• Includes tissue-specific data from 275 breast cancer and 357 prostate cancer transcription factor ChIP-seq tracks.
• Prioritized variant subsets of 1 million, 1.5 million, and 2 million SNVs are provided.

Limitations

• Column-level documentation is absent; field semantics must be inferred after download.
• Row count is unknown, which may limit suitability assessment.
• The dataset is very small at 9.8 KB, indicating a limited scope, likely a list or summary rather than raw genomic data.

Provenance

Source

figshare, author Qing Li.

Collection Method

Generated via a transfer learning framework adapting the Enformer model to tissue-specific ChIP-seq data.

Freshness

Last updated 2026-05-06 18:00:35; freshness should be verified.