The Hox gene family is present in animals across phyla and has the potential to explain major changes of the bauplan (body plan - the overall layout of the body). Here is a link to a good introductory article on the topic: WHERE DO TOES COME FROM? If you would like much more detailed articles on the ties between the Hox gene family and vertebrate limb development and evolution, see Anterior / Posterior Patterning - An Introduction to the Hox Gene Family, Where Did Limbs Come From? (both lectures from a wonderful online course for a Vertebrate Evolution And Development class),   Hoxa-13 and Hoxd-13 play a crucial role in the patterning of the limb autopod and two articles from Pediatric Research -  Homeobox Genes in Embryogenesis and Pathogenesis and  Reconstructing the History of Human Limb Development: Lessons from Birth Defects. The first of these articles is particularly interesting in that it shows several "atavisms" (developmental throwbacks to the body shape of evolutionary ancestors in modern animals) caused by the knocking out or disabling of one or a few Hox genes. For example, knocking out the Hoxa-2 gene resulted in mice with "an atavistic skeletal structure corresponding to the reptilian upper jaw (or pterygoquadrate) cartilage". The article concludes that knocking this gene out causes the mice to develop a jaw similar to that of therapsids - the link between mammals and reptiles. I guess creationists will simply write this off as a great big coincidence, but it is impossible for those who are willing to go where the data leads them to ignore this important link between the fossil record (the evolution of "reptiles" (actually, early amniotes) to therapsids to mammals - accompanied by a major rearangement of the bones of the jaw) and the genome of a modern mammal (shutting off the Hoxa-2 gene leads to development similar to the therapsid jaw in a mouse). This paper also noted that shutting off the Hox-4 genes converted the mouse occipital bones to occipital vertebrae - another presumed atavism as this is the pattern seen in "agnathans" - jawed fishes presumed to be the ancestors of all vertebrates. Another study,  Regulation of number and size of digits by posterior Hox genes: a dose-dependent mechanism with potential evolutionary implications showed that as the amount of HOX gene product present in mouse embryos was gradually reduced, embryo effects led first to ectrodactyly, then to olygodactyly and then to adactyly (no fingers). Interestingly, the researchers saw that on the way from having five fingers to having no fingers in the embryos, there was a step where the embryos were polydactyl - had more than five fingers. Knowing that the first amphibians were polydactyl (see items E. (Ichthyostega) and F. (Acanthostega) in the figure below), the authors proposed the hypothesis that the changes from polydactyl (more than five fingers) limbs in the earliest amphibians to pentadactyl limbs (having five fingers) in the common ancestor of all more recent tetrapods may have been controlled by changes in the regulation of HoxA and HoxD gene expression. Once again we see the possibility of explaining the observed fossil record by observations of the behavior of Hox genes in a modern organism.
 

 Comparison of paired anterior fins of lobe finned fishes (A-D) and limbs of early tetrapods (E, F)  A. Sterropterygion, B. Sauripterus, C. Panderichthys, D. Eusthenopteron, E. Ichthyostega, F. Acanthostega

Original drawing was found here. Here we have a possible explanation for the formation of a new morphological feature - limbs with digits - from the paired fins of sarcopterygian (lobe finned) fishes. Some of these fish (notably Eusthenopteron) have bones in their paired fins that are very similar to the bones of tetrapod limbs. Specifically, they have a single bone (similar to the humerus or femur) followed by paired bones (similar to the radius and ulna or fibula and tibia of tetrapods). Did you know that there were fish with limb bones? I didn't until I learned about Eusthenopteron. Scientists think these fish used their limbs to walk on the sea (or lake or river) bed. What Eusthenopteron lacks are digits - having fin rays instead (although Sauripterus (a very recent discovery) is a closely related fish that does have 8 digits just like the earliest amphibians- see the illustration). However, the fossil record supplies us with examples of tetrapods that are quite similar to Eusthenopteron - Acanthostega and Icthyostega.

The crossopterygian fish Eusthenopteron is linked to the early  amphibian Icthyostega by a number of characteristics: (1)  same pattern of skull bones as Icthyostega, (2) internal nostrils  (found only in land animals and sarcopterygians - a taxonomic  group encompassing lungfish and crossopterygians), (3) teeth  like amphibians', (4) a two-part cranium (icthyostegids are  the only other vertebrates that have this characteristic), and  (5) same vertebral structure. (derived from McGowan, 1984,  152-153)

[The paragraph above was copped from here.]

Anyway, the fossil evidence coincides very nicely with what is currently being learned about the Hox genes. Also, please note that the Hox genes are known to occur in clusters - gene families,  groups of closely related (but not identical genes) that are produced by gene duplications - which have been observed in the laboratory.

For another view on the fish / amphibian link, see Glenn Morton's Transitional Forms: Fish to Amphibian Transition Documented . Some here may be interested to know that Glenn ghost-wrote the anti-evolution section of Josh McDowell's well known Christian apologetics book "Evidence that Demands a Verdict". Glenn has apparently since concluded that ceationism has nothing to offer those who are interested in science.

 The Evolution Evidence Page