I have been told many times that similarity does not imply evolution. At this moment I would like take the time to announce that I resoundingly agree. Many times after I have talked about similarities between creatures that walk this earth to a Creationist only to be told similarity does not imply evolution. Please let it be known that I wholeheartedly agree. If only similarity was observed, that would be insufficient to prove evolution true. I'll let Talk Origins explain some more
Mere similarity between organisms is not enough to support macroevolution; the nested classification pattern produced by a branching evolutionary process, such as common descent, is much more specific than simple similarity. Real world examples that cannot be objectively classified in nested hierarchies are the elementary particles (which are described by quantum chromodynamics), the elements (whose organization is described by quantum mechanics and illustrated by the periodic table), the planets in our Solar System, books in a library, or specially designed objects like buildings, furniture, cars, etc.
It is possible to classify anything into a hierarchical structure, but the traits used will in the end be weighted subjectively. This results in a variety of possible orderings for sets not naturally . Take a set of cars. We could choose a variety of traits to order the nested result--say first by color, then by manufacturer. However , the outcome would be colored by our personal opinion about how to classify them best. In other words no mater what ordering we choose for the cars, any nested hierarchy we choose would be supported by the data. But only certain things can be objectively classified into a unique nested hierarchy.
Because heredity (and occasional horizontal gene transfer in single celled organisms) are the only known ways that two organisms can share proteins, it can be concluded that the sharing of proteins is indicative of them sharing a common ancestor. More often than not, two organisms related by this structure share the same errors in their proteins, because heredity passes those down too. The same is true of any retroviral insertions.
One major example a gene that is broken in humans that controls Vitamin C synthesis. This gene is also broken in our primate cousins, namely Chimpanzees, Bonobos, Gorillas, and Orangutans. Additionally our closest cousin, the Chimpanzee has 48 chromosomes. The gene markers located at the ends of the last two chromosomes in their genes match precisely to the markers located in the middle of our chromosome 46. Furthermore our cytochrome c's differ by only four base pairs, in addition them to having retroviral insertions at the same spots in our DNA. Note that retroviral insertions are rare and occur randomly, so the probability of them occurring in two different species in the same place in their DNA is next to impossible. The only conclusion that can be reasonably drawn is a hereditary relationship. Not to mention we share about 95% of our DNA.
The evolutionary process is a Markov process, which is simply a system that has multiple states, changing from one state to the next with a certain probability. The reason for this has to do with the random probability for mutation, combined with the gene inheritance rules discovered by Mendel. It can be rigorously proven that replicating Markov processes give rise to a nested hierarchy. Therefore a Markov process that reproduces would invariably give rise to a nested tree structure of offspring.
Modern bioinformatics combined with genomic sequencing of a variety of species has allowed us to molecularly reconstruct the tree of life. Based on a set of organisms gene sequences we can determine what is the most probable tree for that set. We can assume heredity based on aforementioned factors; retroviral insertions, errors in the same locations, and similarity if combined with these other factors. When constructed from the genomic and molecular evidences, the phylogeny obtained match precisely with the same tree constructed by morphological and physiological characteristics. This provides independent confirmation of the phylogenetic tree from both molecular and morphological evidences.
To reiterate, this isn't about similarity. This is about the fact that given a set of organisms there is a tree that can be optimally contracted to represent a hereditary relationship between them. So, unlike the set of cars, where the tree would be arbitrarily drawn based upon the color or manufacturer with no bias toward either, these organisms can be grouped into an optimal tree.
