| Literature DB >> 35849028 |
Po Lam Chan1,2,3, Susana Lauw1, Ka Lee Ma1, Nelson Kei1, Ka Leong Ma1,2, Yiu On Wong1,2, Ho Yan Lam2, Yee Yung Ting3, Tsz Kwan Yau1, Wenyan Nong1, Dandan Huang4,5, Yichun Xie1, Peter Chi Keung Cheung1,3, Hoi Shan Kwan1,2,3.
Abstract
The use of probiotics to improve health via the modulation of gut microbiota has gained wide attention. The growing volume of investigations of probiotic microorganisms and commercialized probiotic products has created the need for a database to organize the health-promoting functions driven by probiotics reported in academic articles, clinical trials and patents. We constructed ProBioQuest to collect up-to-date literature related to probiotics from PubMed.gov, ClinicalTrials.gov and PatentsView. More than 2.8 million articles have been collected. Automated information technology-assisted procedures enabled us to collect the data continuously, providing the most up-to-date information. Statistical functions and semantic analyses are provided on the website as an advanced search engine, which contributes to the semantic tool of this database for information search and analyses. The semantic analytical output provides categorized search results and functions to enhance further analysis. A keyword bank is included which can display multiple tables of contents. Users can select keywords from different displayed categories to achieve easily filtered searches. Additional information on the searched items can be browsed via the link-out function. ProBioQuest is not only useful to scientists and health professionals but also to dietary supplement manufacturers and the general public. In this paper, the method we used to build this database-web system is described. Applications of ProBioQuest for several literature-based analyses of probiotics are included as examples of the various uses of this search engine. ProBioQuest can be accessed free of charge at http://kwanlab.bio.cuhk.edu.hk/PBQ/. Database URL: http://kwanlab.bio.cuhk.edu.hk/PBQ/.Entities:
Mesh:
Year: 2022 PMID: 35849028 PMCID: PMC9290863 DOI: 10.1093/database/baac059
Source DB: PubMed Journal: Database (Oxford) ISSN: 1758-0463 Impact factor: 4.462
Figure 1.The workflow of ProbioQuest. The database works with continuous data retrieval from 3 public sources via API function. Our in-house Perl script transforms the raw data into the desired structure and then inputs it to MySQL. Subsequently, Solr creates segments of the content and indexes according to the title and abstract or summary of the processed data. The user interface was built by using JavaScript for searching, with its statistical function and analysis feature.
Figure 3.An overview of the user interface of ProBioQuest. (A) Home page includes a keyword cloud of the most frequent keywords from all seed articles, a search bar which offers suggestions while the user is typing and a help box for users unfamiliar with the system’s details. (B) Autism as an example of input and the search result. Each article is linked out to ClinicalTrials.gov, PubMed.gov or PatentsView with a simple click. (C) Novel feature in Keyword Bank tab. This tab shows relevant keywords in descending order of the number of articles they co-existed, which aids secondary key concept discovery of the interested topic. All keywords are divided into categories. (D) The secondary keyword can be appended into the search bar or linked out. (E) Search result can be exported as CSV format or linked out to search for additional information.
Breakdown by source of entries in the ProbioQuest database (Data accessed on 31 December 2021)
| Number | |
|---|---|
| Total entries in the database | 2 894 703 |
| Entries from Pubmed.gov | 2 656 818 |
| Entries from PatentsView | 32 536 |
| Entries from ClinicalTrials.gov | 205 349 |
Figure 2.Flow chart of semantic analysis and exploration on the user interface of ProBioQuest. Users can simply input the keyword of interest, and ProBioQuest will show related search results. Users can then click the results for further reading or export the selected result. To explore deep searching, ProBioQuest provides a statistical analysis of the search result and Keywords Bank allows further searching by adding a secondary keyword.
Figure 4.The number of articles containing ‘probiotic/probiotics’ from 1950 to 2021 and articles of top 30 Disease/Health in 5-years groups. (A) The number of articles on probiotics has been growing rapidly during the past two decades. (B) Popular topics such as inflammation, diarrhea and cancer maintained a high rate of occurrence in the articles from 1991 to 2021.
Figure 5.2001 to 2020 for 2020 top 10 Disease/Health conditions revealed by ProBioQuest. The top 10 most popular Disease/Health keywords in 2020 were selected for further analysis of their trends during the two decades from 2001 to 2020. (A) Inflammation and dysbiosis have clear trends of marked growth in numbers. (C–D) Specific diseases of awareness on metabolic disorders like obesity, diabetes and oxidative stress and diseases of abnormal cell growth like cancer and necrosis also increased in the past 10 years, although to a lesser extent.
Figure 6.Flowchart of using ProBioQuest to evaluate a probiotic strain-Lactobacillus rhamnosus GG. Searching Lactobacillus rhamnosus GG in ProbioQuest returned 929 results. There were 890 articles from Pudmed.gov. To narrow down the number of results to those of more specific interest, a second keyword (‘Bifidobacterium’ or ‘inflammation’) was added. This reduced the number of results to 108 and 99, respectively. The addition of more keywords would further reduce the number of results and narrow the field of enquiry.