Greetings,

I’m happy to announce support for multiple phenotypes by FaST-LMM’s single_snp function. Running with more phenotypes can generate more data, so we’ve also added new features related to caching and filtering out unneeded output lines.

A new notebook demonstrates all the new features.

A university user requested these feature and then used them to do a multi-day run with 16,000 phenotypes. That run found a couple areas with promising associations.

To install FaST-LMM, start with Miniconda or Anaconda.

      (If already installed,  un-install it with “pip uninstall fastlmm”.) Then …

conda install "mkl==2019.4" "scipy" "numpy"
pip install --no-build-isolation fastlmm

• Carl