BaranziniLab/KG_RAG

Disease_ With_ Relationship_ To_ genes.pickle file

Closed this issue · 5 comments

xuzhaoyang-svg commented

Thanks for your work. What data does the Disease_ With_ Relationship_ To_ genes.pickle file contain? And how to create this file?

karthiksoman commented

Hi @xuzhaoyang-svg
disease_with_relation_to_genes.pickle holds the names of diseases whose vectors are computed and stored in vectorDB. Here, we only chose those disease names that have a relationship with Genes in the biomedical knowledge graph called SPOKE. Hence the name disease_with_relation_to_genes.
This file was created by running a CYPHER query against the SPOKE graph (in neo4j):

MATCH(d:Disease)-[r:ASSOCIATES_DaG]->(g:Gene)
RETURN d.name AS d_name

Let me know if this answers your question.

xuzhaoyang-svg commented

Store the query results obtained from the CYPHER query in . pickle file format by calculating vectors.

karthiksoman commented

It is stored in pickle file format 'for computing vectors' NOT 'by calculating vectors'. Let me know if that clarifies your question.

xuzhaoyang-svg commented

Thank you, I understand now.

karthiksoman commented

Awesome! I am closing this issue then.