In ShawnBrad/SearchPubmed: Interface to Search PubMed

knitr::opts_chunk$set(eval = FALSE, include = TRUE)

Introduction

This package aims to enhance the utility of the DiscoverNature package by searching the pubmed database using the entrez api. The package introduces the ability to perform a wider range of searches which are not limited to a single publisher and also returns abstracts along with detailed author information. The package was design to be simple to use for users, while still making the myriad of search options provided in the entrez api available to more advanced users.

Getting Started

The package can be installed from github using the following commands in R

#library(devtools)
#install_github("ShawnBrad/SearchPubmed", force = T)
library(PubSearch)
library(tidyverse)

Basic literature search

The search.pubmed( ) is used to query pubmed for a topic of interest.

paper.results <- search.pubmed(term = 'cells', retmax = 5)
paper.results

Sorting results

Results from searches can be sorted by date, relevance, title or author using the sort argument.

 paper.results_titlesort <- search.pubmed(term = 'cells', sort = 'title', retmax = 5)
 paper.results_titlesort %>%
   dplyr::select(article.titles,article.date)

 paper.results_datesort <- search.pubmed(term = 'cells', sort = 'most+recent', retmax = 5)
 paper.results_datesort %>%
    dplyr::select(article.titles,article.date)

Restricting the number of records returned

The maximum number of records returned can be limited using retmax argument. there is an upper limit of 100,000)

 paper.results <- search.pubmed(term = 'cells', retmax = 10)
 paper.results

Restricting search by date

Two options are available for restricting searches by date. The 'reldate' argument restricts searches to the past n days, while the dates arguments restricts searches to user provided dates

relative date searches

 paper.results_last10days <- search.pubmed('cells', reldate = '10', retmax = 5)
 paper.results_last10days

provided date searches

 paper.results_june1to30 <- search.pubmed('cells', dates = c('2019/06/01','2019/06/30'), retmax = 10)
 paper.results_june1to30

You can also specify year and/or months via the dates argument

 paper.results_junetojuly <- search.pubmed('cells', dates = c('2019/06','2019/07'), retmax = 5)
 paper.results_junetojuly

 paper.results_2017to2018 <- search.pubmed('cells', dates = c('2017','2018'), retmax = 5)
 paper.results_2017to2018

Advanced searches

Search phrases can be limited to specific fields by specifying the name of the field using the fields argument.

 paper.results_cells <- search.pubmed(term = 'cells',fields = 'Title', retmax = 5)
 paper.results_cells %>%
  pull(article.titles)

The total set of supported fields have been made available with the package.

 data(PubmedTags) 
 PubmedTags

The fields arguments will modify the entire search phrase,however individual search terms can be restricted to speficic fields by providing the field tags along with the term. Journal abbreviations can be found on the NCBI website

 paper.results_cells_nature <- search.pubmed(term = 'cells[TI] nature[TA]', retmax = 5)
 paper.results_cells_nature %>%
   select(article.journal, article.titles)

Using tags and fields arguments simultaniously is not recomended.

Opening articles

We have also included a convinient way to directly open papers from searches

 paper.results <- search.pubmed(term = 'cells', dates = c(2015,2018), retmax = 5)

#open.article(search = 'MMP-1', results = paper.results, index = 'article.titles')

Getting author information

Author information is returned as a tibble contained in the article.authors column of a searched results. To view easily we have provided the see.authors function which takes the same arguments as the open.article function. We reccomend either removing the a

 see.authors(search = 'MMP-1', results = paper.results, index = 'article.titles')

If exporting the saved results to text reccomend either removing the article.authors column column collapsing the authors into a string prior to export

ShawnBrad/SearchPubmed documentation built on Jan. 19, 2020, 5:19 p.m.