Resources

Here are some useful resources for downloading MEDLINE and PubMed Open Access (PubMed OA) XML data.

Links to download PubMed OA and MEDLINE dataset

Below, we provide links for downloading PubMed OA and MEDLINE data

PubMed Open-Access (OA) dataset is available at http://www.ncbi.nlm.nih.gov/pmc/tools/ftp/. Here is the FTP link for downloading the bulk of dataset. In the FTP link, you can go to oa_bulk folder to see the full available tar files.
the MEDLINE XMLs are available here ftp://ftp.nlm.nih.gov/nlmdata/.medleasebaseline/gz/
the MEDLINE XMLs weekly updates are available here ftp://ftp.nlm.nih.gov/nlmdata/.medlease/gz/
MEDLINE Document Type Definitions (DTDs) file is available at this link. We can use it to see available tags from a given MEDLINE XML.

Download PubMed OA figures

Here, we explain how to download PubMed OA figures corresponded to the parsed information from parse_pubmed_caption function

In pubmed_parser, you can use parse_pubmed_caption to parse figures (to be specific figure_id) and captions corresponding to a manuscript.
To download the images corresponding to a given PMC or PMID, you can download a CSV file from ftp://ftp.ncbi.nlm.nih.gov/pub/pmc/oa_file_list.csv first. The file will have columns PMID, Accession ID (PMC), and File. In File column, you can see the path to download a tar file of an XML and corresponding figures in the following format oa_package/08/e0/PMC13900.tar.gz.
You can use the path to download a tar file for a given PMID or PMC in a following format: ftp://ftp.ncbi.nlm.nih.gov/pub/pmc/oa_package/08/e0/PMC13900.tar.gz. If you want to download all the tar files, check out ftp://ftp.ncbi.nlm.nih.gov/pub/pmc/oa_package/ to see all the files.

PMC Copyright Notice

When you use Pubmed Parser to parse information from the website, do not download them as a bulk. Your IP might get banned from doing it. Please see copyright notice when you scrape data from website here.

Alternative implementation of MEDLINE parsers

There are a few implementation to parse MEDLINE dataset. You can see below if you are interested to these alternative implementations.

MEDLINE Kung-Fu which uses medic to parse MEDLINE to database
MEDLINEXMLToJSON implemented in JavaScript