Skip to content
View michael-denyer's full-sized avatar
🧨
🧨

Block or report michael-denyer

Block user

Prevent this user from interacting with your repositories and sending you notifications. Learn more about blocking users.

You must be logged in to block users.

Content in all repositories owned by your account will be closed.
Maximum 250 characters. Please don’t include any personal information such as legal names or email addresses. This note will only be visible to you.
Report abuse

Contact GitHub support about this user’s behavior. Learn more about reporting abuse.

Report abuse
michael-denyer/README.md

Michael Denyer, scientific Python, genomics, data engineering, AI, and repository stars

I build bioinformatics and AI tools for the jobs that eat your afternoon: slow analyses, unfamiliar codebases, and agents that need the same explanation again. Mostly Python and C, with occasional detours into cats and cosmic destruction.

Python Rust C PySpark Databricks Docker

Bioinformatics · AI development · Other apps · Contributions · Website

Bioinformatics

JAMMA

Get more GWAS done with the compute you have. JAMMA accelerates kinship and association testing with native C kernels, processes genotypes in chunks to control memory use, and reuses eigendecompositions across phenotypes. Keep familiar GEMMA commands or use the Python API.

PyPI · Benchmarks · Lean 4 proofs

My numpy-mkl fork builds Intel MKL NumPy and SciPy wheels with 64-bit NumPy indexing for much larger GWAS.

Fast Beagle

Phase and impute larger cohorts with less waiting and less RAM. These C ports of Beagle 5.4 and 5.5 preserve each Java release's VCF output byte for byte at the same thread count, and add BGEN output. The 5.5 benchmarks show 1.5× to over 3× faster runs, with memory use reduced by up to 3.5×.

Beagle 5.5 edition · Beagle 5.4 edition · Benchmarks

pyLocusZoom

Give your GWAS hits some context. Turn result tables into publication-ready figures or interactive plots, with LD, genes, and fine-mapping evidence alongside the association signal. Works with Pandas or PySpark, with built-in dog and cat references and support for other species.

PyPI · Plot gallery

Manhattan and QQ plots produced with pyLocusZoom

Manhattan and QQ plots produced with pyLocusZoom.

AI development

  pstack