mims-harvard/TDC

Suggesting a new data function: Knowledge Graph Mastery

abearab opened this issue · 1 comments

Suggesting a new data function: Knowledge Graph Mastery

In this PR mims-harvard/TDC#207, I started a few features that I believe can help to build, handle, explore, and integrate knowledge graphs based on the final PrimeKG data format (i.e. simply a pandas data frame with same column names). Here I mentioned a bit more details as part of a discussion with @payalchandak, @ayushnoori, and @marinkaz.

The proposed Knowledge Graphs data function can be useful for:

  1. Future updates of PrimeKG
  2. Building new graphs from other resources and datasets, e.g. mims-harvard/PrimeKG#12
  3. Building context-specific knowledge graphs, see "Future directions" in Machine learning predicts cellular response to genetic perturbation. Nat Biotechnol (2023).

Also, I have already used this functionality for other use cases and can provide more updates moving forward.


Note that the data function title suggested by chatGPT haha

image

Originally posted by @abearab in #23 (comment)

Thanks! This looks interesting. I will give some comments on the implementation in the PR!