Codon statistics

Analysis of tlie global statistics of protein sequences has recently allowed light to be shed on anotlier puzzle, tliat of tlie origin of extant sequences [170]. One proposition is tliat proteins evolved from random amino acid chains, which predict tliat tlieir length distribution is a combination of the exponentially distributed random variable giving tlie intervals between start and stop codons, and tlie probability tliat a given sequence can fold up to fonii a compact... 

And finally let us come to the most equilibrated of all codes, the one in which every amino acid is specified by three codons (Figure 5.9). In this case we cannot figure out what kind of world would have been generated, but surely it would have been different from ours, because proteins would have had different statistical mixtures of amino acids and therefore different chemical properties. 

Figure 5.9 An hypothetical genetic code in which all amino acids are codified by three codons. It would be the most balanced code, in the sense that all amino acids would have the same statistical frequency, but it has not been nature s choice.

ANACONDA is a software package specially developed for the study of genes primary structure. It reads ORFeomes downloaded from public databases in FASTA format and uses a set of statistical and visualization methods to reveal information about codon context, codon usage, and nucleotide repeats within ORFs. The general features of ANACONDA are described below ... 

ANACONDA uses contingency tables as the basic statistical methodology and identifies preferred and rejected codon pairs of an ORFeome through the analysis of adjusted residuals values of the contingency tables. The following list highlights the main statistical procedures performed by the software. 

ANACONDA then calculates the value of the Pearson s chi-squared statistic and the adjusted Pearson residual values. Pearson s statistic represents a global measure of the difference between observed and expected codon frequencies (20). 

If the hypothesis of independence between the variables A and B, i.e., between contiguous codons, is rejected ANACONDA determines the contributions of each 64 x 61 codon pairs to Pearson s statistic value computing the adjusted residual values (21). 

Data processing (quantification). The imported sequences are then processed according to the statistical methodology that reveals the irregularities in the codon... 

In contrast to protein-gene finders that are routinely used for genome annotation, noncoding RNA (ncRNA) gene finders are still in their infancy. Systematic de novo prediction of ncRNAs is hindered by the fact that there are no common statistically significant features in primary sequence (e.g., open reading frames or codon bias), which could be exploited for efficient algorithms. 

To avoid implying something that is not in fact true, I repeat a point made briefly in Chapter 2, that mitochondria get many of their proteins from the host cell, and these proteins are coded in the nucleus with the usual universal code. The proteins that are coded by the mitochondrial DNA constitute a small fraction of the proteins that work in the mitochondria.) Second, the small nnmber of mitochondrial genes makes it mnch more likely than it would be for the nuclear DNA that certain codons shonld fall into disuse as a simple statistical fluctuation, thereby facilitating the sort of mechanism I have just discussed. These sorts of considerations suggest that it is quite possible that the mitochondrial variations arose after the universal code became established, thongh they do not, of course, exclude the alternative supposition that they are survivals of extremely ancient versions of the code. 

Since also this "a priori" statement is verified by the Fortran code analysis, it is by no means possible at present to conclude that the complementary strand gene distribution is statistically different from the coding strand gene distribution. Also the fact that the complementary distribution seems not to reach a peak at 140 < L < 150 codons cannot be applied in order to demonstrate a meaningful difference since, as we said previously, 5 out of the 17 genes analysed code for proteins (globin) of very close amino acid number. 

Codons used for the initial gene population and for mutation at each generation of the genetic algorithm are chosen statistically according to the codon table settings. The degree to which the... 

This "batch" sequencing approach can be extended to determine the frequency at which a given amino acid residue appears at a particular sequence position. These statistics are important for analyzing the functional sequences resulting from selection or screening of large combinatorial cassette libraries. Integrated density values should be entered into a vector [d] after normalization for the dideoxy / deoxynucleotide termination statisitics. The first four elements of [d] correspond to the normalized band densities for A, G, C and T in the first codon position, repeated for the... 

---