Posted: Jun 24, 2010 12:19 pm
Jaredennisclark:
I'm not confident enough in my knowledge of the subject to really get into the dirty facts with you, as I'm just learning myself.
CharlieM:
Thanks for your honesty, I'm still learning myself.
Jaredennisclark:
I will tentatively put forward that, from what I understand, a long succession of slight modifications can account for the flagellum.
CharlieM
Not without the system importing ready assembled parts and modifying them to perform a new function. That's why Matkze proposed, "five subsystem-level co-option events"
Take the hook for instance. It is really a marvel of engineering. Over one hundred units of the same protein make up a tubular universal joint. Each of these units are too large to pass through the tube to reach their destination so they do not fold into their final three D shape until they are on site so to speak. And formation can take place as the drive is spinning. The proteins of the hook allow torque to be transmitted through an angle so that the tail is spinning on its axis and not rotating round the axis of the basal body, hence the term, "universal joint". This is not an easy task to achieve from an engineering perspective. The proteins of the hook need to be compressed on the inside of the bend and then expanded at the outside during each rotation of the flagellum. Where do the forces come from to manipulate the structure in such a manner? The proteins have to bind to each other and to their neighbouring proteins in such a way as to achieve this function.
No other system like it has been found so co-option from something similar is not an option. Matkze and Pallen therefore propose that it developed from other homologous proteins and they give a table with these proteins shown. The hook protein consists of approximately 400 amino acids, the sequence (DNA and amino acid) can be found on the 'net. This can be compared with its homologs and I would suggest that a very large number of changes would need to be made to any common ancestral protein to produce the various functions that we see in the homologs. Are chance occurrences of long successions of slight modifications up to this task? Well belief that this is the case becomes a matter of faith, because it hasn't been observed or demonstrated to work as far as I can see.