Some data about this result are given in an article in Science vol. 277 July 11, 1997, pp. 176-178. The method used to extract the DNA was the polymerase chain reaction, which on old and damaged DNA is highly error prone. The DNA was in fragments about 100 base pairs long, which had to be pieced together into a region 379 base pairs long. This process of piecing together took three months of work. Among the fragments found were some that looked like modern human DNA, and these were considered to be contamination.
This process was done twice, and the same sequence was obtained both times. The resulting sequence was compared with 986 distinct sequences from living humans. The sequence differed from these in an average of 25.6 positions. Living humans differ in this region in an average of 8 positions, but the maximum difference is 24 positions. But the pattern of mutations in the Neanderthal sequence was different than in modern humans. For comparison, in this region, there are 55 differences between humans and chimpanzees.
It seems that there is still room for interpretation concerning the accuracy and validity of this result. Even if the Neanderthal sequence is correct, one needs to consider whether these mutations result from evolution or from some environmental factor. Furthermore, there is still considerable controversy about the existence of the missing links. In addition, "some of the Neanderthals get more severe in their archaic morphology as they approach the end of the Neanderthal sequence, the opposite of evolutionary expectations," according to Dr. Marvin L. Lubenow, creationist author of Bones of Contention. This would seem to indicate that the Neanderthals were degenerating and not evolving to become more fit. This could indicate that these mutations were the result of some environmental factor.
Another point to consider is the implication of the mutation rates assumed for human evolution. Typical rates of mutation assumed for human ancestors are about one substitution per base pair every 200 million years. (See Mol Biol Evol 1985 March; 2(2):150-174.) (Actually, many different figures are used, and the real rates of mutation are unknown. This rate is based on assumed evolutionary time scales.) Assuming 10 percent of the human genome codes for protein, this means about 300 million base pairs code for protein. This would give on the average 300 million substitutions every 200 million years. Probably 2/3 of these would influence the protein coded, and 1/3 would be silent, leading to 200 million harmful mutations in 200 million years, about one a year. With a generation time of 30 years, this is an average of 30 harmful mutations per offspring for humans from each parent, or 60 in all. The chance of an offspring being free from harmful mutations is then 1/(2.718) 60 , which is astronomically small. This would imply that the human race is rapidly degenerating. Some of these mutations might be recessive, but in 10,000 generations or so, their frequency should increase enough so that this effect should be realized. In fact, when the population reaches equilibrium, we should expect only about one in 2.718 60 fertilized eggs to survive, at this mutation rate. A justification for this calculation method as well as other calculations which are problematical to the theory of evolution can be found here (near the end) and here.