Genomic machine learning

Inspiration

Many scientists and researchers could benefit from machine learning techniques, but don't know how to write software. I like scientific programming.

What it does

Takes two datafiles, a genome and your own data, and puts both into a machine learning algorithm(not sure what yet). It gives you back any interesting relationships between the two files. Ideally the user doesn't have to 'clean' their data first, or even think about it.

How I built it

1. Get python FTP client to get genomic data
2. take 2 files
3. put them into target machine learning algo
4. output results ## Challenges I ran into
5. deciding what algorithms to use
6. restricting choice of genomic data, as file size can be large ## Accomplishments that I'm proud of

What I learned

What's next for Genomic machine learning

• make a pretty GUI
• cross-platform or single platform only?
• give a choice of algorithms to use
• determine if python is a suitable choice

Built With

• python
• qt
• scikit-learn