skihobbes said:
Hey all--I am a high school physics teacher, A student asked me a question about the fact that DNA does not increase the amount of information in the strand--the information is just reshuffled. I have searched and I have read many people saying this is inaccurate but nobody gives any specific examples. Can anyone help me out with this argument?
NobodySpecial said:
Depends on the strict definition of information.
There is no more information in the works of Shakespeare than there is an a random sequence of letters of the same length.
This, along with arguments about increasing complexity and the 2nd law of thermodynamics, are common attacks on evolution.
skihobbes said:
Going back to the book analogy mentioned above--let's say you have 1000 letters. Has the number of letters actually increased. For example---1000 letters become 1100. I explained that the dictionary has more information than the same number of random letters--and he understood that. So I guess it isn't "information" but quantity of letters increasing.
Barakn said:
Polyploidy does indeed increase the amount of DNA in a cell (and can act as a speciation event), but does that really count as an increase in information? Would you know more about the Bill of Rights if I were to hand you two copies of it rather than just one? And does horizontal transfer of a drug-resistance plasmid explain the origin of the information in that plasmid?
nobahar said:
Hi, Barakn; thanks for the reply to my post.
I think polyploidy offers the opportunity for an increase in information.
There's a lot of misconceptions in this thread regarding information theory mixed in with a lot of good information regarding genetics.
First- the information content of a string of symbols. The information content can be uniquely and precisely specified in two ways: 1) the Shannon entropy and 2) the Kolmogorov complexity.
The Shannon entropy is a measure that is appropriate for the transmission of a stream of signals, and relates to how well you can *predict* the next symbol, given that you know the value of the current symbol. For a random string of binary digits, each bit is associated with k ln(2) units of entropy (kT ln(2) units of energy). English text carries approximately 12 bits of entropy per word.
http://people.seas.harvard.edu/~jones/cscie129/papers/stanford_info_paper/entropy_of_english_9.htm
Kolmogorov complexity relates to the inherent information content of what is encoded, and is more difficult to quantify. One way is to set the Kolmogorov information as the minimum length message required to specify the object, or alternatively, the number of computer instructions required to reproduce the result:
http://en.wikipedia.org/wiki/Kolmogorov_complexity
The use of 'information' and 'entropy' in this context are opposite our intuitive sense of the terms- a random string of bits has maximal information, but makes no sense to us. So, it is common to use 'negentropy' of a string in order to make the concepts intuitive.
So now the question becomes, "How does DNA encode information?" I'm not an expert, but there are several ways to formulate a response:
1) each base pair of DNA corresponds to (IIRC) k ln(4) bits of information. So increasing the length of DNA increases the information.
2) Not all DNA encodes 'useful' information: introns and 'junk' DNA (to be sure, much of what we used to think of as noncoding DNA is used for alternative purposes). This is at the limit of my understanding of information theory, as the Kolmogorov measure is more appropriate.
3) One gene can encode multiple proteins through splicing and post-translational modification. This is at the limit of my understanding of genetics.
4) cells and higher-level structures operate using regulatory feedback networks: for example, the jak/stat pathway leads to nuclear translocation of a DNA-binding protein to regulate gene expression. It's not clear how this higher-level network relates to the information content of the genome.
5) Our cells contain more DNA than what's in the nucleus:
mitochondrial DNA, for example. And AFAIK, it's not clear how centrosomes multiply after the cell has divided.
So, bottom line, DNA does encode information in a quantifiable way.
There's another problem pertaining to dealing with ID arguments- your opponent is not constrained by logic. No amount of logical argument will suffice to convince them that life and evolution are subject to the same physical constraints as say, a block of wood sliding down an inclined plane.