Source code for the @JC_pathogenomic (https://twitter.com/JC_pathogenomic)
1. Clone the repo:
$ git clone https://github.com/happykhan/journalbot.git
2. Install Python packages:
$ python -m pip install -r requirements.txt
3. Set environment variables for Twitter authorisation (see https://apps.twitter.com/):
Create a new twitter app, and generate a consumer_key, consumer_secret, access_token, access_token_secret. Export these as Environment Variables.
Also create an NCBI/Pubmed Entrez email (https://www.ncbi.nlm.nih.gov/account/?back_url=https%3A%2F%2Fwww.ncbi.nlm.nih.gov%2Fgquery%2F).
E.g. in a script:
export consumer_key='xxxxxxxxxxxxxxxxxxxx'
export consumer_secret='xxxxxxxxxxxxxxxxxxxx'
export access_token='xxxxxxxxxxxxxxxxxxxxx'
export access_token_secret='xxxxxxxxxxxxxxxxxxx'
export entrez_email='user@email.com'
Then source this file to create the variables:
$ source /path/to/credentials_file
The bot will search for queries contained within the config/seartchterms.txt file. These should be composed in PubMed query syntax
An example file would look like (use boolean operators and parentheses to evaluate more complex searches):
Smith J
Smith J[author]
Jones K
E coli
polymer AND synthesis
nucleus NOT atomic
human[organism] AND topoisomerase[protein name]
Smith J OR Jones K
Smith John M [FAU]
Installation is easy with a Miniconda/Anaconda install:
$ wget https://repo.continuum.io/miniconda/Miniconda2-latest-Linux-x86_64.sh && bash Miniconda2-latest-Linux-x86_64.sh
Follow installation prompts. Allow appending to $PATH in .bashrc.
$ which python # should yield: ~/anaconda2/bin/python or similar
should return a path including miniconda. If not, it's using the wrong python binary still, and you need to alter the order of paths in $PATH.