search

greenelab / snorkeling

Extracting biomedical relationships from literature with Snorkel 🏊
Other
59 stars 17 forks source link

Planning the Snorkel for whether a compound treats a disease #1

Open dhimmel opened 7 years ago

dhimmel commented 7 years ago

@danich1 is a new rotation student in the Greene Lab :champagne:. We were thinking a good rotation project would be to extract medical indications from the literature. The project is intended as a pilot. The ultimate goal is to automate the integration of all biological knowledge into a single hetnet.

We were thinking Compound–treats–Disease relationships were a good place to start for the following reasons:

  1. Comprehensive catalogs of treatments will be crucial for computational drug repurposing approaches.
  2. In Project Rephetio, we physician curated a gold standard catalog of indications, called PharmacotherapyDB. In addition, we have cataloged treatments under investigation in clinical trial. Both of these catalogs of treatments could be used to create labeling functions.
  3. We shouldn't have to invent tagging methods for diseases and compounds, as there should already be mature/implemented solutions.

Project Rephetio used Disease Ontology as its disease vocabulary and DrugBank as its drug vocabulary. But we're flexible here.

Paging @ajratner and @stephenbach -- creators of Snorkel.

dhimmel commented 7 years ago

@danich1, for guidance check out the following (private communication by @ajratner):

If interested, you can also see the chemical-disease relation task I was mentioning. We're still cleaning it up, adding explanatory comments, etc--it's not in tutorial form yet--but it's the most up-to-date application we currently have posted online: https://github.com/HazyResearch/snorkel/tree/cdr-tutorial/tutorials/cdr

ajratner commented 7 years ago

Seems like a good task to me! Further discussion around (a) chemical vs. "compound" and (b) initial relation type on other threads

dhimmel commented 7 years ago

I met @thodrek (Theo Rekatsinas) at the Moore Early Career Symposium in Hawaii. He's part of HazyResearch and helped me understand snorkel. We also snorkeled in real life too and saw a :turtle:.

I learned that snorkel will extract sentences that either support a positive (+1) or negative (-1) relationship. However, we will have to do an additional step to "fuse sentences" into a single judgement or score on whether a given relationship exists. @thodrek's paper on SLiMFast is a good place to look for inspiration (blog post).

@thodrek, thanks for the help. Let us know whether you're interesting in contributing to this project and we can add you to the repo. Contributions could be giving feedback on GitHub issues, pull request review as requested, or anything else valuable. Cheers.