Fig. 7: Keyword co-occurrence networks in CHO (blue), HEK293 (purple), and combined (green) cell line datasets.

A co-occurrence network consists of edges and nodes where each note represents an author keyword, and the edges show how often those keywords reported to co-occur in the same publication. Thicker edges are indicative of stronger co-occurrence. The networks are represented in the following order: The co-occurrence network for CHO cell author keywords in publications retrieved using the bioprocess query (A) and without the bioprocess query (B) the co-occurrence network for HEK293 cells with the bioprocess query clearly separated into two clusters (C) and without the bioprocess query in a single connected network (D), the absence of a co-occurrence network for both cell line author keywords due to limited number of author keywords in the publications of this dataset (n = 2) (E), and the co-occurrence network for both cell line author keywords without the bioprocess query, showing two distinct networks (F).