Sequential pattern mining for discovering gene interactions and their contextual information from biomedical texts

Table 5 Example of information for a sequential pattern

PMID	Gene 1	Gene 2	BioGRID verdict	Sentence
10204582	SHC1	CRKL	not in BioGRID	These results suggest a fundamental role for the
				tyrosine phosphorylation of Cbl, CrkL, SLP-76, and
				<gene_name="SHC1" > and the association
				of Cbl with <gene_name="CRKL" >, SLP-76,
				and Nck in Fc gammaRI signaling in human
				macrophages.
10204582	CBL	CRKL	in BioGRID	PP1, a specific inhibitor of Src kinases, inhibited
				the Fc gammaRI-induced respiratory burst,
				as well as the tyrosine phosphorylation of
				<gene_name="CBL" > and its inducible
				association with <gene_name="CRKL" >.

Table 5 gives 2 interactions (highlighted in bold in the table) detected by the sequential pattern: “AGENE association with AGENE”.
The meaning of the columns is: the id number of the paper in PubMed, the genes that interact, the verdict of the oracle and the sentence where the interaction is recognized. For instance, the pattern detects that in paper 10204582, an interaction between genes SHC1 and CRKL is mentioned, because the pattern matches a sentence of the abstract of the paper, but according to BioGRID, there is no interaction between SHC1 and CRKL and the discovered interaction in the sentences in paper 10204582 is unexpected because not in BioGRID and interesting. The pattern also detects that in this paper, an interaction between genes CBL and CRKL is mentioned, and indeed, according to BioGRID, there is an interaction between CBL and CRKL mentioned in paper 10204582.

ISSN: 2041-1480