N.Wells
Posts: 1836 Joined: Oct. 2005
|
Quote (GaryGaulin @ Nov. 22 2016,16:13) | Quote (N.Wells @ Nov. 22 2016,07:12) | Why do you think that working with all possible triplets is better than just working with the sequences taken one nucleotide at a time? |
The banding method was in part inspired by Guenter's info. This is from program comments:
Quote | From: http://www.basic.northwestern.edu/g-buehler/genomes/genome.htm
The universal Codon spectrum (majorityn distribution)- also the work of countless inversions and inverted transpositions. The infinity of conceivable Codon distributions. Chargaff's second parity rule specifies only that the numbers of Codons are equal to the numbers of their reverse complements in every sufficiently long genomic DNA strand. But it does not specify the numbers of each Codon. In fact, there is an infinity of possible Codon distributions, that all fulfill the rule and, yet, are all different from each other. This is easy to show. There are 64 different [three letter] Codons. They can be divided into 2 groups where the members of one group are the reverse complements of the members of the other. For example, one such division could consist of the following groups. GROUP 1: AGT, ATT, CAT, CCT, CGG, CGT, CTG, CTT, GAA, GAG, GAT, GCA, GCG, GCT, GGA, GGC, GGG, GGT, GTA, GTC, GTG, GTT, TAG, TAT, TCG, TCT, TGA, TGG, TGT, TTA, TTG, TTT. GROUP 2: ACT, AAT, ATG, AGG, CCG, ACG, CAG, AAG, TTC, CTC, ATC, TGC, CGC, AGC, TCC, GCC, CCC, ACC, TAC, GAC, CAC, AAC, CTA, ATA, CGA, AGA, TCA, CCA, ACA, TAA, CAA, AAA
-----------
The illustration banding pattern shows both the GROUP 1 and GROUP 2 abundances. On the very left you will see the corresponding three letter "AGT, ATT, CAT,," pair being shown in that row. Their unusual closeness made it possible to show 64 triplet possibilities in only 32 comparisons. |
After finding out how well this works for mathematically dissecting chromosomes the next logical step became to use the same idea for other sequence lengths. The program started with three, but all other lengths of unique coding are now needed. How that all together branches out helps show the systematics of the system needing to be modeled. |
I think you are misunderstanding Albrecht-Buehler's point. Insertions, deletions, translocations / inversions, recombinations and so on are indeed the source of variation, to the extent that a huge number of possible combinations and permutations of every strand of DNA get tried, although the vast majority fail horribly without even getting out of the gate. The fact that reverse complements are as common as the original sequences indeed testifies to the anarchy and brutality of the process, and the thoroughness of the exploration of possibilities. However, I don't see any point to comparing species by assessing all possible triplets and reverse complements - even with alternative splicing, only a tiny set of all possibilities are utilized in any given species, and considering the rest only obscures relevant similarities.
Compare "illumination" and "Elimination": Method A: Illumination E_limination One has an i instead of an e, an extra l, and a u instead of an i, and everything else is the same, giving a 9/12 similarity.
Method B: Compare: ill, llu, lum, umi, min, ina, nat, ati, tio, ion, mlr, lrg, rgz, gzm, zmr, mrn, rnf, nfo, foo, & oor, versus eli, lim, imi, min, ina, nat, ati, tio, ion, mlr*, lrg, rgz, gzm, zmr, mrn, rnr, nro, rov, for 12 / 20 matches
* Using noitanimulli and noitanimile as reversals, and thus mlrgzmrnrov and mlrgzmrnfoor as complements (ab becomes yz, etc.)
You seem to be pursuing something much closer to Method B. Method B shows the variety of possibilities that can be explored by mutation, but how is this helpful in evaluating DNA similarities between species? Comparison of sequences seems best served by analyzing unit-by-unit match-ups.
Evolution works because mutation proposes promiscuously, while selection disposes rigorously. As Albrecht-Buehler notes, life reproduces fast enough to avoid total DNA randomization, aided by the evolution of DNA repair mechanisms and the fact that once a stretch of DNA does something functional, different versions of it, its reverse complement, and grafting it into other stretches of DNA are also likely to do something interesting, albeit slightly better, slightly worse, or in a radically different context. Creationists don't get the way evolution works when they insist that all mutations must be bad, and you don't get that when you rail cluelessly against natural selection.
|