Amino Acid Compositions of Proteins
Functions to identify proteins, get and set amino acid compositions, and calculate thermodynamic properties from group additivity.
1 2 3 4 5 6 7
character, name of protein; numeric, indices of proteins (rownumbers of
character, name of organism
logical, compute per-residue counts?
data frame, amino acid composition in the format of
character, physical state
character, protein sequence
numeric, abundances of proteins
logical, return the weighted average of amino acid counts?
character, path to file with amino acid compositions
A protein in CHNOSZ is defined by a name and by the counts of amino acids, stored in
thermo$protein. The purpose of the functions described here is to identify proteins and work with their amino acid compositions. From the amino acid compositions, the thermodynamic properties of the proteins can be estimated (Dick et al., 2006) for use in other functions in the package.
Often, the names of proteins are sufficient to set up calculations using functions such as
species. The names of proteins in CHNOSZ are distinguished from those of other chemical species by having an underscore character ("_") that separates two identifiers, referred to as the
organism (but any other meaning can be attached to these names). An example is LYSC_CHICK.
The first few functions provide low-level operations:
iprotein returns the rownumber(s) of
thermo$protein that match the protein names. The names can be supplied in the single
protein argument or as separated
organisms (without the underscore). Any protein not matched returns an NA and generates a message.
ip2aa returns the row(s) of
thermo$protein that match the supplied protein names, OR the protein indices found by
residue to TRUE to return the per-residue composition (i.e. amino acid composition of the protein divided by total number of residues). For this function only, if the
protein argument is a data frame, it is returned unchanged, except for possibly the per-residue calculation.
aa2eos calculates the thermodynamic properties and equations-of-state parameters for the completely nonionized proteins using group additivity with parameters taken from Dick et al., 2006 (aqueous proteins) and LaRowe and Dick, 2012 (crystalline proteins and revised aqueous methionine sidechain group). The return value is a data frame in the same format as
state indicates the physical state for the parameters used in the calculation (aq or cr).
The remaining functions are more likely to be called directly by the user:
seq2aa returns a data frame of amino acid composition, in the format of
thermo$protein, corresponding to the provided
sequence. Here, the
protein argument indicates the name of the protein with an underscore (e.g. LYSC_CHICK).
aasum returns a data frame representing the sum of amino acid compositions in the rows of the input
aa data frame. The amino acid compositions are multiplied by the indicated
abundance; that argument is recycled to match the number of rows of
average is TRUE the final sum is divided by the number of input compositions. The name used in the output is taken from the first row of
aa or from
organism if they are specified. This function is useful for calculating bulk amino acid compositions in stress response experiments or localization studies; see
read.expr for examples of its use.
read.aa returns a data frame of amino acid composition based on the contents of the indicated
file, which should be a CSV file with the same column names as
add.protein completes the loop; any amino acid composition returned by the
*aa functions described above can be added to
thermo$protein using this function to be made available to other functions in the package.
The amino acid compositions of proteins in
aa with the same name as one in
thermo$protein are replaced.
The value returned by this function is the rownumbers of
thermo$protein that are added and/or replaced.
Dick, J. M., LaRowe, D. E. and Helgeson, H. C. (2006) Temperature, pressure, and electrochemical constraints on protein speciation: Group additivity calculation of the standard molal thermodynamic properties of ionized unfolded proteins. Biogeosciences 3, 311–336. http://dx.doi.org/10.5194/bg-3-311-2006
uniprot.aa for getting amino acid compositions from a FASTA file or downloading them from UniProt.
protein.info for higher-level functions (chemical formulas, summaries of reaction coefficients and energies),
more.aa for getting amino acid compositions for model organisms from additional data files in the
extdata/protein directory, and
read.expr for working with protein abundance and subcellular localization data.
Examples of stability calculations for proteins are in
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52
# search by name in thermo$protein ip1 <- iprotein("LYSC_CHICK") ip2 <- iprotein("LYSC", "CHICK") # these are the same stopifnot(all.equal(ip1, ip2)) # two organisms with the same protein name ip3 <- iprotein("MYG", c("HORSE", "PHYCA")) # their amino acid compositions ip2aa(ip3) # their thermodynamic properties by group additivity aa2eos(ip2aa(ip3)) # an example of an unrecognized protein name ip4 <- iprotein("MYGPHYCA") stopifnot(is.na(ip4)) # manually adding a new protein # Human Gastric juice peptide 1 aa <- seq2aa("GAJU_HUMAN", "LAAGKVEDSD") ip <- add.protein(aa) stopifnot(protein.length(ip)==10) # the chemical formula of this peptide stopifnot(as.chemical.formula(protein.formula(ip))=="C41H69N11O18") # we can also calculate a formula without using add.protein as.chemical.formula(protein.formula(seq2aa("pentapeptide_test", "ANLSG"))) # read a fasta file, calculate H/C and O/C for all proteins # and averages by polypeptide chain, residue and mass ffile <- system.file("extdata/fasta/HTCC1062.faa.xz", package="CHNOSZ") aa <- read.fasta(ffile) pf <- as.data.frame(protein.formula(aa)) plot(pf$H/pf$C, pf$O/pf$C, pch=NA) points(pf$H/pf$C, pf$O/pf$C, pch=20, cex=0.5, col="grey") # average composition per polypeptide chain chainaa <- aasum(aa, average=TRUE) chainpf <- as.data.frame(protein.formula(chainaa)) points(chainpf$H/chainpf$C, chainpf$O/chainpf$C, pch=15, col="red") # average by amino acid residue pl <- protein.length(aa) resaa <- aasum(aa, abundance=pl, average=TRUE) respf <- as.data.frame(protein.formula(resaa)) points(respf$H/respf$C, respf$O/respf$C, pch=16, col="red") # average by mass pm <- mass(pf) massaa <- aasum(aa, abundance=pm, average=TRUE) masspf <- as.data.frame(protein.formula(massaa)) points(masspf$H/masspf$C, masspf$O/masspf$C, pch=17, col="red") # add a legend and title legend("topright", pch=c(20, 15, 16, 17), col=c("grey", rep("red", 3)), legend=c("protein", "chain average", "residue average", "mass average")) title(main=paste("O/C vs H/C for HTCC1062 proteins\n", "n =", nrow(aa)))