Predict RNA-seq, CAGE, DNase, and ChIP tracks from DNA sequences.
npx skills add https://github.com/PKU-YuanGroup/OpenAI4S --skill borzoi-pku-yuangroup
Please help me install this Agent Skill. Skill: borzoi Source: https://github.com/PKU-YuanGroup/OpenAI4S/tree/main/skills/borzoi Command: npx skills add https://github.com/PKU-YuanGroup/OpenAI4S --skill borzoi-pku-yuangroup
Predicts genome-wide functional tracks from DNA sequences for use cases like scoring the regulatory effect of variants on expression/accessibility, generating predicted coverage tracks for a locus, and prioritizing non-coding variants.
Predict RNA-seq functional tracks for the locus using Borzoi:
from borzoi_pytorch import Borzoi
model = Borzoi.from_pretrained("johahi/borzoi-replicate-0").cuda().eval()
High-intent search queries and answers about installing and using this skill.
To predict RNA-seq functional tracks from DNA sequences, you can use Borzoi by loading the pre-trained model via PyTorch and passing your genomic locus data directly to the model for evaluation.
You score the regulatory effect of non-coding variants by using DNA sequence predictions to calculate track deltas, prioritizing variants based on their predicted impact on expression and accessibility.
Yes, genome-wide functional track prediction supports CAGE, DNase, and ChIP tracks simultaneously, allowing you to analyze predicted coverage across multiple genomic regions for comprehensive regulatory analysis.
Yes, GPU computation is required for DNA sequence processing. The model is designed to be loaded onto a CUDA-enabled GPU using PyTorch to handle genome-wide functional track predictions efficiently.
The best way to analyze a specific locus is to generate predicted coverage tracks from the DNA sequence, then use these predictions to score and prioritize non-coding variants within that region.
Variant prioritization using predicted track deltas relies entirely on DNA sequence inputs, meaning predictions are constrained to supported genomic regions and require GPU resources for execution.
ovachieverProvides Python-based bioinformatics tools for analyzing DNA, RNA, and protein sequences.
JuanMS20Deploy high-throughput LLM serving endpoints with vLLM and OpenAI-compatible APIs.
RamanEbrahimiAutomate high-throughput sequencing data analysis with deepTools workflows.
pradyumnasagarPredict transcription factor binding and variant effects from genomic sequences.