pubmedBatch in Python

Send your searches in batch against Pubmed!

Installation

You will need python, flask, mongodb to make the app work

And to speed up the app, you should download the following dtd files

wget -O ~/.biopython/Bio/Entrez/DTDs/nlmmedlinecitationset_160101.dtd http://www.ncbi.nlm.nih.gov/corehtml/query/DTD/nlmmedlinecitationset_160101.dtd
wget -O ~/.biopython/Bio/Entrez/DTDs/pubmed_160101.dtd http://www.ncbi.nlm.nih.gov/corehtml/query/DTD/pubmed_160101.dtd
wget -O ~/.biopython/Bio/Entrez/DTDs/bookdoc_160101.dtd http://www.ncbi.nlm.nih.gov/corehtml/query/DTD/bookdoc_160101.dtd
wget -O ~/.biopython/Bio/Entrez/DTDs/efetch.dtd http://eutils.ncbi.nlm.nih.gov/eutils/dtd/20131226/efetch.dtd

Configuration

The configuration is set in config.cfg. You can change the settings according to your use. By default, the app is running on localhost:8000/pubmedbatch

Introduction

if you have a list of genes to search against Pubmed, you can put them into a column together with other relevant information, and upload it as a csv file. Note that the csv file requires a header.

The app will try to calculate a Pubmed score based on the occurrence of the OR terms in the titles/abstracts, and sort articles / genes based on the score per article and score per gene, respectively.

It recognises some column headers, such as 'Func', 'ExonicFunc', 'Pred' in order to heuristically aggregate the predicting damaging effect of a given variant:

If splic | stop | frame | del | insert found, set Pred score to 1000
Else, each D or A gives 10, P gives 5, C gives 6, T | B | N gives -1

It also recognises column with header 'FILTER'. If 'FILTER' is PASS, the del row button is coloured green. If it is FAIL, the del row button is coloured yellow. Else the button is coloured red.

Features

It saves the results of each search in a database. It won't search the same term in Pubmed again until 30 days later (can be changed in config.cfg). After 30 days, it will only update the search results by searching Pubmed for articles published in the recent 30 days.
The table is sortable by any column.
The Pubmed ID is clickable.
You can provide a gene list (space separated) to highlight genes.
You can provide a blacklist to skip genes.
You can use the files in the testfile folder to play with the app.
One caveat of the app is, deleting row won't delete it in the actual data.

Future work

Will take tab delimited files

logust79 / pubmedPython