Pubmed mining for occupational idiopathic pulmonary fibrosis papers

Background: Optimal strategies to extract relevant content from Pubmed for occupational lung disease topics are not clearly defined and the utility of mining techniques has not been assessed. Aims and Objectives: To evaluate, given a ‘seed’ of known relevant papers (here 12 occupational IPF case-control studies), the use of relative citation ratio (RCR)[1], jaccard similarity (JS)[2] and coverage(C) to identify additional relevant papers. Methods: We used the python programming language and the Pubmed application programming interface to extract all papers citing the 9seed9 papers and to calculate and sort by RCR, JS, and C. Full analysis online.[3] Results: Papers citing 9seed9 papers sorted by RCR, JS, and C. Table 1 and Figure Conclusions: Pubmed mining techniques can help to identify additional relevant papers. Refs: 1. http://journals.plos.org/plosbiology/article?id=10.1371/journal.pbio.1002541 2. https://en.wikipedia.org/wiki/Jaccard_index 3. https://github.com/drcjar/litsearch/blob/master/litsearch.ipynb