# Download our tables from NCBI's FTP site. Accessed 14:30PST, 18 December 2012
prok <- read.table("ftp://ftp.ncbi.nlm.nih.gov/genomes/GENOME_REPORTS/prokaryotes.txt", sep="\t", comment.char="!", header=T)

# Pull release dates, while dropping rows lacking a release date.
prok  <- as.Date(prok$Release.Date[prok$Release.Date != '-'],format="%Y/%m/%d")

# Bin our dates by month and year, tabulate, and save to a dataframe.
prok.cut <- as.data.frame( 
  table(
    as.Date(
      cut(prok, "month")
      )
    )
  )

# Correct our column titles, calculate a running total, and reconvert from factor to date
colnames(prok.cut) <- c("Date", "Total")
prok.cut$Total <- cumsum(prok.cut$Total)
prok.cut$Date <- as.Date(prok.cut$Date)


# Draw our plot
library("ggplot2")
p <- ggplot(prok.cut, aes(Date, Total))


# Save our plot to SVG
library(grDevices)
svg(filename='ncbi-genomes.svg', width = 15, height = 5)
p +  geom_area() +  xlab("") +  ylab("")
# Alternately, with titles
# +  ggtitle("Bacterial and archeal genome sequences submitted to Genbank") + xlab('Time') +  ylab("Total number of genomes")
dev.off()