Finding literature
The library
Find, star, select, organise, read and export the papers of a project.
One table for every paper
The Library lists every paper in the project with where it has got to: screening, full text, extraction and review. Decisions are made on the stage pages; the Library is for finding papers and organising your reading.
The search box looks in titles, journals, DOIs, author names and the notes people wrote. What you type is matched as written. Views such as "Needs PDF" or "Verified" filter the table without changing any paper, and Columns lets you hide the columns you do not need.
Stars, collections, tags and notes
A star is your own bookmark: only you see your stars, and every member can use them. Collections and tags are shared with the project and are only for organising; nothing scientific reads them. Notes are kept with the paper and can be searched.
Selecting papers
Tick papers to work with several at once (up to 50). The selection stays while you move between pages of the table in the same browser tab. With papers selected you can star them, export them, run Find full text for them from the Full text page, or ask the assistant to "summarise the selected papers".
Reading and attachments
A paper with a PDF has a Read link that opens it in the reader. The Full text column shows the paper’s full-text status in the same words as the Full text page. If an attached file can no longer be opened, the row says so. A paper recorded as a preprint is marked, and when the library also holds its published version, the row links to it.
The details of a row also show which manuscripts cite the paper, the databases it was found in, and links to look it up through your institution.
Exporting
Export the papers of the current view, or only the selected ones, as RIS, BibTeX, CSV or CSL-JSON. The file holds bibliographic details only. Exporting is limited to the owner and to members the owner has allowed to export, and each export is recorded in the project’s activity. A view that is too large to export in one file is refused with the limit stated; it is never cut short without saying so.
Duplicates and corrections
Two records of the same paper can be merged from a row’s details after you have seen a preview of what will be kept. Nothing is merged without your confirmation. "Refresh metadata" re-reads a paper’s details from the databases and shows the current and proposed values side by side before anything is saved.
Related
Searching the literature
Which databases are searched for you, which you search yourself, and what each search records.
Importing reference files
Bringing in papers by DOI or PubMed ID, from a pasted reference list, or from RIS, BibTeX, CSV and Excel files.
Removing duplicates
How duplicate papers are found, and why every merge waits for your confirmation.
Getting full text, legally
How full text is found automatically, what the platform will not fetch, and how the manual list works.
Reading, searching and annotating a PDF
Open a paper beside its details, find words in it, highlight passages and keep notes.