Darwin Help

(1) It assigns a Dayhoff matrix computed at PAM distance 250 to the global variable DM.

(2) It computes 1266 Dayhoff matrices for various PAM distances between 0.049 and 1000 and assigns the list of such matrices to the global variable DMS.

(3) It computes the amino acid natural frequencies and assigns them to the global variable AF.

(4) It assigns the global variable logPAM1 with the logarithm of the mutation matrix (at PAM distance 1) being used.

By default, with no arguments, it uses the data derived from the entire SwissProt database in Nov 1991 (Benner, Gonnet and Cohen). This can be altered in two ways:

(a) by assigning the global variable NewLogPAM1 with the logarithm of a PAM 1 mutation matrix, all the computations will be based on this mutation matrix.

(b) by passing a count matrix as argument all the computations will be based on this count matrix. A count matrix has the counts of mutations (and non mutation on the diagonal) for a large sample of alignments. Normally if two amino acids X and Y are aligned, we will add 1/2 to Counts[X,Y] and 1/2 to Counts[Y,X].

(c) by calling the function with a rate matrix and a frequency vector, the computations will be based on these parameters

If the counts are only on the amino acids A, C, G and T, (and the rest of the counts are just 1 on the diagonal and 0 elsewhere), the Dayhoff matrices produced are suitable to align DNA sequences. Actually this is the standard and simplest way of aligning DNA sequences. The system knows about the following count matrices, which can be used as argument of CreateDayMatrices: