JAX

Use when working with DeepRVAT — a deep learning framework for rare variant association testing (RVAT) that learns gene-level embeddings from rare variant annotations using an auto-encoder, then tests

2 shared topics • 2 shared operations

Geneformer — transformer-based foundation model pretrained on ~30 million single-cell transcriptomes for context-specific gene network analysis. Supports fine-tuning for cell type classification, gene

2 shared topics • 2 shared operations

Use when working with Sei — the FunctionLab sequence-to-activity deep learning framework for predicting chromatin profiles and regulatory activity from DNA sequences. Sei takes 4096 bp DNA windows and

2 shared topics • 2 shared operations

Evo — genomic foundation model for DNA sequence modeling at single-nucleotide resolution. Uses StripedHyena architecture (7B parameters) trained on 2.7M prokaryotic and phage genomes (OpenGenome datas

2 shared topics • 1 shared operation

GATK VQSR (Variant Quality Score Recalibration) — machine learning-based variant filtering for GATK Best Practices germline pipelines. Trains a Gaussian mixture model on truth/training resource datase

2 shared topics • 1 shared operation