Period (days) needed for a manuscript to be accepted after receievd in journals of the area of bioinformatics

bioinformatics_journal_period.R

#! /usr/bin/Rscript

# This script is used for getting the period (days) needed for a manuscript to be 
# accepted after receievd in journals of the area of bioinformatics.

# ref: http://www.r-bloggers.com/bioinformatics-journals-time-from-submission-to-acceptance-revisited/
    
    
################################################################################
# Some bioinformatics journals
titles <- c("Bioinformatics", "BMC Bioinformatics", "BMC Genomics", 
	"PLoS Computational Biology")

################################################################################
# Get XML files from pubmed
getJournalXML <- function(title) {
  library(rentrez)
  print (title)
  term <- paste(title, "[JOUR]", sep = "")
  term <- paste(term, 'AND ("2013/01/01"[Date - Publication] : "3000"[Date - Publication])')
  e <- entrez_search("pubmed", term, usehistory = "y")
  f <- entrez_fetch("pubmed", WebEnv = e$WebEnv, query_key = e$QueryKey, 
  	rettype = "xml", retmax = e$count)
  d <- xmlTreeParse(f, useInternalNodes = TRUE)
  outfile <- paste(gsub(" ", "_", title), "xml", sep = ".")
  saveXML(xmlRoot(d), outfile)
}

 
# saves XML files in current working directory
sapply(titles, function(x) getJournalXML(x))

################################################################################
# period
xml2day <- function(xmlfile){
    library(XML)
    journal_xml <- xmlParse(xmlfile)
    
    
    ################################################################################
    received_year <- xpathSApply(journal_xml, "//PubmedData/History", function(x) {
        if (xpathSApply(x, "boolean(./PubMedPubDate[@PubStatus='received']/Year)")) {
            xpathSApply(x, "./PubMedPubDate[@PubStatus='received']/Year", xmlValue)
        } else {
            NA
        }
    })
    
    received_month <- xpathSApply(journal_xml, "//PubmedData/History", function(x) {
        if (xpathSApply(x, "boolean(./PubMedPubDate[@PubStatus='received']/Month)")) {
            xpathSApply(x, "./PubMedPubDate[@PubStatus='received']/Month", xmlValue)
        } else {
            NA
        }
    })
    
    received_month <- sapply(received_month, function(x) {
        if (is.na(x)) return (NA) 
        else if (nchar(x)!=2) return (paste0(0,x)) 
        else return (x) 
    }
    )
    
    received_day <- xpathSApply(journal_xml, "//PubmedData/History", function(x) {
        if (xpathSApply(x, "boolean(./PubMedPubDate[@PubStatus='received']/Day)")) {
            xpathSApply(x, "./PubMedPubDate[@PubStatus='received']/Day", xmlValue)
        } else {
            NA
        }
    })
    
    received_day <- sapply(received_day, function(x) {
        if (is.na(x)) return (NA) 
        else if (nchar(x)!=2) return (paste0(0,x)) 
        else return (x) 
    }
    )
    
    ################################################################################
    accepted_year <- xpathSApply(journal_xml, "//PubmedData/History", function(x) {
        if (xpathSApply(x, "boolean(./PubMedPubDate[@PubStatus='accepted']/Year)")) {
            xpathSApply(x, "./PubMedPubDate[@PubStatus='accepted']/Year", xmlValue)
        } else {
            NA
        }
    })
    
    accepted_month <- xpathSApply(journal_xml, "//PubmedData/History", function(x) {
        if (xpathSApply(x, "boolean(./PubMedPubDate[@PubStatus='accepted']/Month)")) {
            xpathSApply(x, "./PubMedPubDate[@PubStatus='accepted']/Month", xmlValue)
        } else {
            NA
        }
    })
    
    accepted_month <- sapply(accepted_month, function(x) {
        if (is.na(x)) return (NA) 
        else if (nchar(x)!=2) return (paste0(0,x)) 
        else return (x) 
    }
    )
    
    accepted_day <- xpathSApply(journal_xml, "//PubmedData/History", function(x) {
        if (xpathSApply(x, "boolean(./PubMedPubDate[@PubStatus='accepted']/Day)")) {
            xpathSApply(x, "./PubMedPubDate[@PubStatus='accepted']/Day", xmlValue)
        } else {
            NA
        }
    })
    
    accepted_day <- sapply(accepted_day, function(x) {
        if (is.na(x)) return (NA) 
        else if (nchar(x)!=2) return (paste0(0,x)) 
        else return (x) 
    }
    )
    
    # filter NA
    ind <- which(!is.na(accepted_day))
    ind1 <- which(!is.na(received_day))
    ind <- intersect(ind, ind1)
    received_year1 <- received_year[ind]
    received_month1 <- received_month[ind]
    received_day1 <- received_day[ind]
    
    accepted_year1 <- accepted_year[ind]
    accepted_month1 <- accepted_month[ind]
    accepted_day1 <- accepted_day[ind]
    
    
    days <- as.numeric(as.Date(paste(accepted_year1, accepted_month1, accepted_day1, sep="/") ) -
                           as.Date(paste(received_year1, received_month1, received_day1, sep="/") ) )
    
    return (days)
    
}


xml_files <- paste(gsub(" ", "_", titles), "xml", sep = ".")
journal_days <- {}
for (i in 1:length(xml_files)) {
    xmlfile <- xml_files[i]
    journal <- titles[i]
    journal_days[[journal]] <- xml2day(xmlfile)
}

################################################################################
# Visualization
boxplot(journal_days, notch=T, outline=FALSE)