I was looking at the examples in Y0, and I came across this one:
sars_large_example = Example( name="SARS-CoV-2 Graph", reference="Jeremy Zucker, Sara Mohammad-Taheri, Kaushal Paneri, Somya Bhargava, Pallavi Kolambkar" ", Craig Bakker, Jeremy Teuton, Charles Tapley Hoyt, Kristie Oxford, Robert Ness, and Olga Vitek. 2021." "Leveraging Structured Biological Knowledge for Counterfactual Inference: a Case Study of Viral Pathogenesis" "- IEEE Journals & Magazine. IEEE Transactions on Big Data (January 2021).", graph=NxMixedGraph.from_str_edges( nodes=[ "SARS_COV2", "ACE2", "Ang", "AGTR1", "ADAM17", "Toci", "Sil6r", "EGF", "TNF", "EGFR", "PRR", "NFKB", "IL6STAT3", "IL6AMP", "cytok", "Gefi", ], directed=[ ("SARS_COV2", "ACE2"), ("ACE2", "Ang"), ("Ang", "AGTR1"), ("AGTR1", "ADAM17"), ("ADAM17", "EGF"), ("ADAM17", "TNF"), ("ADAM17", "Sil6r"), ("SARS_COV2", "PRR"), ("PRR", "NFKB"), ("EGFR", "NFKB"), ("TNF", "NFKB"), ("Sil6r", "IL6STAT3"), ("Toci", "Sil6r"), ("NFKB", "IL6AMP"), ("IL6AMP", "cytok"), ("IL6STAT3", "IL6AMP"), ("EGF", "EGFR"), ("Gefi", "EGFR"), ], undirected=[ ("SARS_COV2", "Ang"), ("ADAM17", "Sil6r"), ("PRR", "NFKB"), ("EGF", "EGFR"), ("EGFR", "TNF"), ("EGFR", "IL6STAT3"), ], ), example_queries=[ Query.from_str(treatments="Sil6r", outcomes="cytok"), Query.from_str(treatments="EGFR", outcomes="cytok"), ], )
However, in the paper referenced, a version of this graph without bi-directed edges is present. The one that includes the bi-directed edges is this reference:
Mohammad-Taheri, S., Zucker, J., Hoyt, C. T., Sachs, K., Tewari, V., Ness, R., & Vitek, O. (2022). Do-calculus enables estimation of causal effects in partially observed biomolecular pathways. Bioinformatics, 38(Supplement_1), i350-i358.
I suggest to create 2 networks. One without any bi-directed edges and one with bi-directed edges.
I was looking at the examples in Y0, and I came across this one:
sars_large_example = Example( name="SARS-CoV-2 Graph", reference="Jeremy Zucker, Sara Mohammad-Taheri, Kaushal Paneri, Somya Bhargava, Pallavi Kolambkar" ", Craig Bakker, Jeremy Teuton, Charles Tapley Hoyt, Kristie Oxford, Robert Ness, and Olga Vitek. 2021." "Leveraging Structured Biological Knowledge for Counterfactual Inference: a Case Study of Viral Pathogenesis" "- IEEE Journals & Magazine. IEEE Transactions on Big Data (January 2021).", graph=NxMixedGraph.from_str_edges( nodes=[ "SARS_COV2", "ACE2", "Ang", "AGTR1", "ADAM17", "Toci", "Sil6r", "EGF", "TNF", "EGFR", "PRR", "NFKB", "IL6STAT3", "IL6AMP", "cytok", "Gefi", ], directed=[ ("SARS_COV2", "ACE2"), ("ACE2", "Ang"), ("Ang", "AGTR1"), ("AGTR1", "ADAM17"), ("ADAM17", "EGF"), ("ADAM17", "TNF"), ("ADAM17", "Sil6r"), ("SARS_COV2", "PRR"), ("PRR", "NFKB"), ("EGFR", "NFKB"), ("TNF", "NFKB"), ("Sil6r", "IL6STAT3"), ("Toci", "Sil6r"), ("NFKB", "IL6AMP"), ("IL6AMP", "cytok"), ("IL6STAT3", "IL6AMP"), ("EGF", "EGFR"), ("Gefi", "EGFR"), ], undirected=[ ("SARS_COV2", "Ang"), ("ADAM17", "Sil6r"), ("PRR", "NFKB"), ("EGF", "EGFR"), ("EGFR", "TNF"), ("EGFR", "IL6STAT3"), ], ), example_queries=[ Query.from_str(treatments="Sil6r", outcomes="cytok"), Query.from_str(treatments="EGFR", outcomes="cytok"), ], )
However, in the paper referenced, a version of this graph without bi-directed edges is present. The one that includes the bi-directed edges is this reference:
Mohammad-Taheri, S., Zucker, J., Hoyt, C. T., Sachs, K., Tewari, V., Ness, R., & Vitek, O. (2022). Do-calculus enables estimation of causal effects in partially observed biomolecular pathways. Bioinformatics, 38(Supplement_1), i350-i358.
I suggest to create 2 networks. One without any bi-directed edges and one with bi-directed edges.