Skip to main content
. 2020 Oct 1;10:16191. doi: 10.1038/s41598-020-73105-0

Figure 1.

Figure 1

Overview of data collection process. Starting from PubMed, we retrieved a list of paper titles and abstracts using the Pubmed Entrez API, and then applied text matching to automatically filtered the search results, obtaining a subset of papers, which were then read and manually evaluated. If papers contained information on the chemical content of cocoa or garlic, we manually extracted the relevant information. Finally, we converted values in comparable units. The “Output” bar shows the result of each step for garlic and cocoa.