Understanding in vivo Models of Depression: A Systematic Review - Records of Full Search

This Zenodo record outlines the full list of journal articles retrieved from the search string as well as a a subset of articles that have been screened by two independent human reviews and reconciled by a third independent screener. 

We carried out a search of 2 online databases for studies reporting animal models of depression. This search, carried out in May 2016, identified 70,365 unique publications (File: Depression-Dataset-SLIM-AllRecords.txt )

Two independent investigators have, to date, screened 5749 of these publications for inclusion or exclusion and these publications form the dataset for this study (File: Updated-training-data.txt  ). 

Several text-mining approaches will be developed for this depression literature search using the results from manual screening to train the machine, where machine-learning software set rules to automatically define each publication as included or excluded without the need for human screening. In this project, we seek to identify the best performing machine learning algorithm for this depression literature search. Performance is measured on sensitivity, specificity, and precision. 

Column Names in Datasets: 

Depression-Dataset-SLIM-AllRecords.txt  -  ID, Author, Year, Title, Journal, Volume, Issue, Pages, Abstract, URL, SetNumber

Depression-Dataset-SLIM-DevelopmentTrainingSet.txt  -  ID, Author, Year, Title, Journal, Volume, Issue, Pages, Abstract, URL, Incl(1)/Excl(0)

Updated-training-data.txt  -  ID, Author, Year, Title, Journal, Volume, Issue, Pages, Abstract, URL, Incl(1)/Excl(0)