Our primary source of information is PubMed. We use the summaries, full text articles (some from behind paywalls).

Additional source of data include:

• Department of Energy Joint Genome Institute in collaboration with the user community.
• National Center for Biotechnology Information Search database
• United States Department of Agriculture – Agricultural Research Service
• EU’s Phenol-Explorer
• GENERATIVE BIOINFORMATICS FROM PETER D’ADAMO

PubMed Literature Review Method

Pubmed has over 33,450,000 citations. We progressively scan these for any citations containing the taxonomical name of bacteria. This results in 1,642,000 candidate articles.

We now examine these articles and filter them to articles containing either a medical condition we are interested in, or a microbiome modifier. This reduces the candidate articles to approximately 100,000 articles.

The next phrase is either manual review of the summary, or thru further textual analysis (for example, excluding articles with “sewers”). The result is around 3100 studies with relevant information (“facts”) which is encoded into our database. One article may produce a single fact; in other cases, 5,000 facts.

The results of the above sources is

• over 1,000,000 Modifier to Bacteria
• over 4,500 Bacteria to Medical Conditions
• over 5,300,000 Bacteria to Enzymes, End Products

Our Goal

The goal is to encode the above facts into a database that can be applied to a microbiome with the use of Artificial Intelligence. Key tables in the database can be exposed via API or replication