arvkevi / clinvar-kaggle

Scripts used to generate the ClinVar conflicting classifications dataset on Kaggle

Hello, 👋 my name is Kevin Arvai. I'm a data scientist with 10+ years of experience in the genomics field.

Connect 🤝
Since you're here, let me know you stopped by. Share your Python or data science story with me on Twitter or LinkedIn. I love hearing about what people are working on in the open-source community!

Favorite project 🧬
I wrote an app that predicts users' ancestry from their genetic data.

Non-GitHub stuff 💻
I like machine learning, open-source software/data, and genomics.
My Real Python articles, blog posts, and Kaggle profile.

clinvar-kaggle's Issues

Data Description

Can I know the features description since most of it abbreviations? what does it refer to, if there is a link of the data description please share it

License File Missing

Hello Kevin,

My name is Ryan Ulaszek and I work at Amazon as a Solution Architect in Life Sciences. Would you be willing to add a license file to the repo? It looks like the Kaggle project is under Create Commons but this repo doesn't have the Creative Commons license file. We think the work you have done is a great example of machine learning in genomics and would like to build an AWS Solution based on it. We will of course point back to your GitHub repo and Kaggle site to give you credit. We would like to use your dataset and the process_clinvar.py file.

Thanks,

Ryan

Add additional features from VEP

An easy enhancement to the dataset would be to add additional features from VEP annotations.

Link to VEP annotations on the ClinVar .vcf

I'm trying to get annotations with one consequence per variant (e.g. one transcript/variant). It will be easier to parse the CSQ field. But at first glance, it doesn't appear to be performing how I expect. Maybe @CarlosBorroto you've run into this before?

Vawk and Gawk dependencies

Hello Kevin,

Is there an easy way to break the dependency on Vawk and Gawk? We can easily add zips of python packages to Glue jobs but binaries are a challenge and Gawk is a binary. Can you think of an easy way to rework the process_clinvar.py script with only python dependencies?

Thanks,

Ryan