seqinr (version 1.0-1)

chargaff: Base composition in ssDNA for 7 bacterial DNA

Description

Long before the genomic era, it was possible to get some data for the global composition of single-stranded DNA chromosomes by direct chemical analyses. These data are from Chargaff's lab and give the base composition of the L (Ligth) strand for 7 bacterial chromosomes.

Usage

data(chargaff)

source

Rudner, R., Karkas, J.D., Chargaff, E. (1968) Separation of B. subtilis DNA into complementary strands, III. Direct Analysis. Proceedings of the National Academy of Sciences of the United States of America, 60:921-922. Rudner, R., Karkas, J.D., Chargaff, E. (1969) Separation of microbial deoxyribonucleic acids into complementary strands. Proceedings of the National Academy of Sciences of the United States of America, 63:152-159.

Details

Data are from Table 2 in Rudner et al. (1969) for the L-strand. Data for Bacillus subtilis were taken from a previous paper: Rudner et al. (1968). This is in fact the average value observed for two different strains of B. subtilis: strain W23 and strain Mu8u5u16. Denaturated chromosomes can be separated by a technique of intermitent gradient elution from a column of methylated albumin kieselguhr (MAK), into two fractions, designated, by virtue of their buoyant densities, as L (light) and H (heavy). The fractions can be hydrolyzed and subjected to chromatography to determined their global base composition. The surprising result is that we have almost exactly A=T and C=G in single stranded-DNAs. The second paragraph page 157 in Rudner et al. (1969) says: "Our previous work on the complementary strands of B. subtilis DNA suggested an additional, entirely unexpected regularity, namely, the equality in either strand of 6-amino and 6-keto nucleotides ( A + C = G + T). This relationship, which would normally have been regarded merely as the consequence of base-pairing in DNA duplex and would not have been predicted as a likely property of a single strand, is shown here to apply to all strand specimens isolated from denaturated DNA of the AT type (Table 2, preps. 1-4). It cannot yet be said to be established for the DNA specimens from the equimolar and GC types (nos. 5-7)."

References

Try example(chargaff) to mimic figure page 17 in http://pbil.univ-lyon1.fr/members/lobry/articles/HDR.pdf. The red areas correspond to non-allowed values beause the sum of the four bases frequencies cannot exceed 100%. The white areas correspond to possible values (more exactly to the projection from R^4 to the corresponding R^2 planes of the region of allowed values). The blue lines correspond to the very small subset of allowed values for which we have in addition PR2 state, that is [A]=[T] and [C]=[G]. Remember, these data are for ssDNA !

To have an overview of the seqinR's functionnality, please consult this vignette: Charif, D., Lobry, J.R. (2005) SeqinR: a contributed package to the R project for statistical computing devoted to biological sequences retrieval and analysis. Springer Verlag, Biological and Medical Physics/Biomedical Series, in preparation.

Examples

data(chargaff)
op <- par(no.readonly = TRUE)
par(mfrow=c(4,4))
xlim <- c(0,100)
ylim <- xlim
par(mai=rep(0,4))
par(c(0.01, 0.99, 0.01, 0.99))
par(xaxs="i")
par(yaxs="i")

for( i in 1:4 )
{
  for( j in 1:4 )
  {
    if( i == j )
    {
      plot(chargaff[,i], chargaff[,j],t="n", xlim=xlim, ylim=ylim,
      xlab="", ylab="", xaxt="n", yaxt="n")
      polygon(x=c(0,0,100,100),y=c(0,100,100,0), col="lightgrey")
      for( k in seq(0,100,by=10) )
      {
        lseg <- 3
        segments(k,0,k,lseg)
        segments(k,100-lseg,k,100)
        segments(0,k,lseg,k)
        segments(100-lseg,k,100,k)
      }
      string <- paste(names(chargaff)[i],"",xlim[1],"% -",xlim[2],"%")
      text(x=mean(xlim),y=mean(ylim), string, cex = 1.5)
    }
    else
    {
      plot(chargaff[,i], chargaff[,j], pch=20, xlim=xlim, ylim=ylim,
      xlab="",ylab="", xaxt="n", yaxt="n")
      iname <- names(chargaff)[i]
      jname <- names(chargaff)[j]
      direct <- function() segments(0,0,50,50, col="blue")
      invers <- function() segments(0,50,50,0, col="blue")
      PR2 <- function()
      {
        if( iname == "[A]" & jname == "[T]" ) { direct(); return() }
        if( iname == "[T]" & jname == "[A]" ) { direct(); return() }
        if( iname == "[C]" & jname == "[G]" ) { direct(); return() }
        if( iname == "[G]" & jname == "[C]" ) { direct(); return() }
        invers()
      }
      PR2()
      polygon(x=c(0,100,100), y=c(100,100,0), col="lightpink")
      polygon(x=c(0,0,100), y=c(0,100,0))
    }
  }
}
# Clean up
par(op)