What does Bioinformatics have to do with Molecular Evolution? 

Problem: Application of first principles does not (yet) work: 

The following chain although (believed to be) mainly determined by the DNA sequence (plus other components of the cell which in turn are encoded by other parts of the genome) can at present not be simulated in a computer.  

DNA sequence ->
transcription ->
translation ->
protein folding ->
protein function (catalytic and other properties) ->
properties of the organism(s) ->
ecology (taking also the non biological environment into account) ->

... .

 

Most scientists believe that the principle of reductionism (plus new laws and relations emerging on each level) is true for this chain; however, this is clearly "in principle" only.
Biology relies on this sequence to work more or less unambiguously (prions), but:

At several steps along the way from DNA to function our understanding of the chemical and physical processes involved is so incomplete that prediction of protein function based on only a single DNA sequence is at present impossible (at least for a protein of reasonable size).

Solution:
Use evolutionary context:

"Nothing in biology makes sense except in the light of evolution"

Theodosius Dobzhansky


Present day proteins evolved through substitution and selection from ancestral proteins. Related proteins have similar sequence AND similar structure AND similar function.

In the above mantra "similar function" can refer to:

  • identical function,

    •
  • similar function, e.g.:
    • identical reactions catalyzed in different organisms; or
    • same catalytic mechanism but different substrate (malic and lactic acid dehydrogenases);
    • similar subunits and domains that are brought together through a (hypothetical) process called domain shuffling, e.g. nucleotide binding domains in hexokinase, myosin, HSP70, and ATPsynthases.

The Size of Protein Sequence Space (back of the envelope calculation):

Consider a protein of 600 amino acids.
Assume that for every position there could be any of the twenty possible amino acid.
Then the total number of possibilities is 20 choices for the first position times 20 for the second position times 20 to the third .... = 20 to the 600 = 4*10^780 different proteins possible with lengths of 600 amino acids.

For comparison the universe contains only about 10^89 protons and has an age of about 5*10^17 seconds or 5*10^29 picoseconds.

If every proton in the universe were a computer that explored one possible protein sequence per picosecond, we only would have explored 5*10^118 sequences, i.e. a negligible fraction of the possible sequences with length 600 (one in about 10^662).

The following is based on observation and not on an a priori truth:

If two proteins (not necessarily true for nucleotide sequences, and for sequences of low complexity) show significant similarity in their primary sequence, they have shared ancestry, and probably similar function.
(although some proteins acquired radically new functional assignments, lysozyme -> lense crystalline). 


To date there is no example known where convergent evolution has let to significant similarity of the primary sequence (although here are examples where similar selection pressures have resulted in similar convergent substitutions in homologous proteins).

THE REVERSE IS NOT TRUE:

PROTEINS WITH THE SAME OR SIMILAR FUNCTION DO NOT ALWAYS SHOW SIGNIFICANT SEQUENCE SIMILARITY
for one of two reasons:

a)  they evolved independently
(e.g. different types of nucleotide binding sites);

or

b)   they underwent so many substitution events that there is no readily detectable similarity remaining.

In particular, PROTEINS WITH SHARED ANCESTRY DO NOT ALWAYS SHOW SIGNIFICANT SIMILARITY
(reason: see B above); many recent advances concern the improved detection of similarity.

 

If you can demonstrate significant similarity using randomization , your sequences are homologous (i.e. related by common ancestry).  Convergent evolution has not been shown to lead to sequence similarities detectable by these means (see above - this might not be true for scores in PSI-blast)

Sorry for request/reminder emails - my spam filters are too eager.

Output from old PRSS version is here.

Table with PRSS results is here.

An approach similar to PRSS is used in the FASTA database search. If one chooses to display a histogram of the search, the output includes the histogram of all the alignment scores obtained with the individual sequences contained in the database. Includes are the actual sequence scores, and the ones that are expected based on a probability distribution. An example is here.

Summary of Terminology:

E-values give the expected number of matches with an alignment score this good or better due to chance alone (no shared ancestry, no cnvergent evolution)

P-values give the probability of to find a match of this quality or better due to chance alone (no shared ancestry, no convergent evolution).

P values are [0,1], E-values are [0,infinity).

Both P and E values should take the size of the databank into consideration, and you should consider to correct for multiple searches to avoid "fishing expedititons".
For small values E=P

z-values give the distance between the actual alignment score and the mean of the scores for the randomized sequences expressed as multiples of the standard deviation calculated for the randomized scores.
For example: a z-value of 3 means that the actual alignment score is 3 standard deviations better than the average for the randomized sequences. Z-values > 3 are usually considered as suggestive of homology, z-values > 5 are considered as sufficient demonstration. (see the but below). A somewhat readable description of E, P, HSP and other values is here.

BUT:
Failure to detect significant similarity does only shows our inability to detect homology, it does not prove that the sequences are not homologous.

Examples:

Jim Knox (MCB-UConn) has studied many proteins involved in bacterial cell wall biosynthesis and antibiotic binding, synthesis or destruction. Many of these proteins have identical 3-D structure, and therefore can be assumed to be homologous, however, the above tests fail to detect this homologies. (for example, enzymes with GRASP nucleotide binding sites are depicted here.)

DNA replication involves many different enzymes. Some of the proteins do the same thing in bacteria, archaea and eukaryotes; they have similar 3-D structures (e.g.: sliding clamp, E. coli dnaN and eukaryotic PCNA, see Edgell and Doolittle, Cell 89, 995-998), but again, the above tests fail to detect homology.

Helicase and F1-ATPase. Both form hexamers with something rotating in the middle (either the gamma subunit or the DNA; D. Crampton, pers. communication). The monomers have the same type of nucleotide binding fold (picture)

Powerpoint Slides on homology and protein space are here

If time discuss exponential functions? (Figs. 1, 2, 3) (More data at the GOLD database here)

  • Exponential growth, decay?
  • log vs ln?
  • dn/dt = k?

If plenty of time go over Kezdy Swinebourne plot here

Assignments:

For Wednesday Sept. 13: Start work on the take-home quiz.
   Remember, next Wednesday is the last class before the quiz is due!

Think about the following question:
If protein space is so big, how come that complex functional molecules were assembled?
(If the answer is not obvious, listen to the AMNH round table discussion on Carl Sagan. Link)

For Friday:

  • Reminder: You should have explored GenBank formatted sequence files. An annotated sample is here. The important features are cross linked to explanations. (Other formats are described here)
  • What is the differences between a locus name, an accession number and the GI number..
  • Read Chapter 4 of the textbook (we return to chapter 3 next week)
  • Glance through the blast query and normal BLAST search tutorials (NOTE: the latter has red arrows at the bottom of the first and second page, that links to the next page!!!)