How To Choose The Right Way Of Data Mining From Pubmed?

Question

[xiaoronglv]

GoPubMed is a famous third party PubMed search engine. User can filter their search result by

• country
• List item
• keywords
• MeSH Tree
• research

To develop this web application, they choose what method,e-utility or load a local medline ?

Answer 1

[Istvan Albert]

It seems that this service works by making use of the public data dumps from NCBI that are then parsed, analyzed and categorized into their own data models and databases and then queried with their own proprietary tools and methods.

[Sudeep]

I also think that they have a regularly updated local abstract repository and searching in GOPubMed retrieves large document sets (more than 1M documents for "DNA") in short response time, which might not be entirely possbile using e-utiilies.

[pradeep]

I agree with Sundeep adn Istvan Albert. However, once bulk of information is processed on data dumps, for daily updates using eutils should be practical.

Answer 2

[Ashutosh Pandey]

I use eutils a lot for my work.

http://www.ncbi.nlm.nih.gov/books/NBK25501/

They are pretty flexible and easy to use and you can do almost anything using them. The result or output generated for a query is in XML format and you can write your own parser or used XML parser in BioPython or BioPerl.