Comparison of missing data handling methods in building variant pathogenicity metapredictors

Description

This project will compare methods for handling missing data in variant annotations for the purpose of building variant pathogenicity metapredictors.

See the research plan for a more detailed description.

To generate a PDF version with a working references section:

$ git clone https://github.com/blueprint-genetics/amiss.git
$ cd amiss/docs/research_plan
$ make

Pandoc and a LaTeX distribution must be installed.

Open science

This project conforms to the principles of open science:

• Open data:
  • We use and reference publically available datasets and will citably archive any data and code we produce using Zenodo with a DOI
• Open source:
  • The source code is freely available under the MIT license at GitHub
• Open notebook:
  • You can follow development from the start on GitHub at https://github.com/blueprint-genetics/amiss
  • The research plan is available in the GitHub repository (see section above for producing a PDF version)
• Open access:
  • We will upload a preprint of the resulting paper(s) on biorXiv
  • We will submit the results for publication in a peer-reviewed open access journal
• Open communication:
  • We intend to present the results in public scientific conferences
• Open collaboration:
  • We welcome
    • ideas, bug reports and comments
    • code contributions through GitHub pull requests