The Journal of Heredity 2001:92(6)
© 2001 The American Genetic Association 92:481-489
Genomic Divergence Between Human and Chimpanzee Estimated from Large-Scale Alignments of Genomic Sequences
From the Department of Life Science, National Tsing Hua University, Taiwan (Chen and Tzeng), and Department of Ecology and Evolution (Chen, Wang, and Li) and Committee on Genetics (Vallender), University of Chicago, Chicago, Illinois.
Address correspondence to Wen-Hsiung Li, Department of Ecology and Evolution, University of Chicago, 1101 East 57th St., Chicago, IL 60637, or e-mail: whli{at}uchicago.edu.
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionConclusionReferences |
|---|
To study the genomic divergence between human and chimpanzee, large-scale genomic sequence alignments were performed. The genomic sequences of human and chimpanzee were first masked with the RepeatMasker and the repeats were excluded before alignments. The repeats were then reinserted into the alignments of nonrepetitive segments and entire sequences were aligned again. A total of 2.3 million base pairs (Mb) of genomic sequences, including repeats, were aligned and the average nucleotide divergence was estimated to be 1.22%. The Jukes–Cantor (JC) distances (nucleotide divergences) in nonrepetitive (1.44 Mb) and repetitive sequences (0.86 Mb) are 1.14% and 1.34%, respectively, suggesting a slightly higher average rate in repetitive sequences. Annotated coding and noncoding regions of homologous chimpanzee genes were also retrieved from GenBank and compared. The average synonymous and nonsynonymous divergences in 88 coding genes are 1.48% and 0.55%, respectively. The JC distances in intron, 5' flanking, 3' flanking, promoter, and pseudogene regions are 1.47%, 1.41%, 1.68%, 0.75%, and 1.39%, respectively. It is not clear why the genetic distances in most of these regions are somewhat higher than those in genomic sequences. One possible explanation is that some of the genes may be located in regions with higher mutation rates.
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionConclusionReferences |
|---|
The genetic divergence between human and chimpanzee has long been an issue of wide interest. A comparison of homologous sequences between the two species not only helps to elucidate the evolutionary history of the human species but also has important implications for the critical changes that separate human from chimpanzee. Prior studies between human and chimpanzee have compared mitochondrial sequences (e.g., Horai et al. 1995), sex chromosome sequences (Bohossian et al. 2000; Kaessman et al. 1999; Nachman and Crowell 2000), and autosomal sequences (Chen and Li 2001). For example, Chen and Li (2001) compared the sequences of short noncoding, autosomal genomic segments between humans and great apes, concluding that the genetic distance between human and chimpanzee is approximately 1.2%. This figure is significantly smaller than those obtained in earlier studies (Bailey et al. 1991; Sibley and Ahlquist 1987). Nevertheless, these studies included only relatively short sequences and limited numbers of genes.
In light of the near completion of the human genome and the rapid accumulation of chimpanzee sequence data, it is desirable to perform a large-scale comparison between the sequences of the two species and reexamine the accuracy of previous estimates. In addition, previous studies did not include repetitive elements in genetic distance calculations. Repetitive elements are estimated to compose 40–50% of the human genome (Gu et al. 2000; Li et al. 2001; Smit 1999) and humans are known to have Alu insertions that are not present in apes (Arcot et al. 1997; Batzer et al. 1994). Further, it has been proposed that repetitive elements may play an important role in human evolution (Rowold and Herrera 2000; Szmulewicz et al. 1998). It is of interest to understand how repetitive elements may affect sequence alignment and whether they display different evolutionary rates than nonrepetitive sequences. Also the overall genetic distance, the distance that takes into account both repetitive and nonrepetitive sequences, can be obtained only when homologous repeats between human and chimpanzee are correctly aligned.
The comparison of human and chimpanzee genomic sequences also will provide a "standard" reference between the two species. This distance can serve as a basis for future biological and medical studies. Genes with substitution rates significantly higher than that of the genomewide average may indicate a higher mutation pressure or positive selection. On the other hand, genes with significantly lower substitution rates may be important for maintaining normal physiological functions. These genes are of great interest to medical, biotechnological, and evolutionary research.
In this study 2.3 million base pairs (Mb) of human and chimpanzee homologous genomic sequences and the coding and noncoding regions of 121 genes were analyzed. The Jukes–Cantor (JC) distances (for genomic and noncoding sequences), the number of nucleotide substitutions per synonymous site (KS), and the number of nucleotide substitutions per nonsynonymous site (KA) (for coding sequences) were derived.
Another focus of this study was the development of a strategy for large-scale sequence alignments that include repetitive elements. Large-scale alignments of genomic sequences, even for closely related species such as human and chimpanzee, pose unexpected difficulties. Insertions and deletions and the presence of repetitive elements complicate the alignment procedure. A method employed to overcome these difficulties is described in this article; the computer program is available from W.-H. Li.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionConclusionReferences |
|---|
Data Retrieval
Human and chimpanzee homologues of annotated genes and chimpanzee unannotated genomic sequences were retrieved from GenBank. The latter were then masked using RepeatMasker (courtesy of the Bioinformatics Group, Institute for Systems Biology, Seattle, WA, and the University of Washington). Six to ten 1–2 kilobase (kb) segments of nonrepetitive sequences of the chimpanzee genome were then extracted for BLASTing against the human genome to identify human counterparts. The human contig hit by all of the chosen nonrepetitive segments of one chimpanzee contig with greater than 95% sequence similarity was determined to be the human counterpart of the chimpanzee contig under study and was retrieved for alignment.
Sequence Alignment and Distance Calculation
All the human and chimpanzee contigs retrieved from GenBank were masked with the RepeatMasker program and the repeats excised from the contig sequences. The resulting sequences were submitted to a computer program developed in our laboratory (unpublished) to determine "anchors" (highly similar sequence segments between two species). These anchors were then used to divide each contig into short segments for alignment with the Clustal W program (Thompson et al. 1994). Repeats were then reinserted between the aligned segments with reference to the alignment coordinates of the nonrepetitive sequences. The repeats were then aligned using the Clustal W program and used as new anchors to adjust the previous alignments of the nonrepetitive segments. Overlapping sequences were excluded with reference to the alignment coordinates. The repetitive and nonrepetitive sequence alignments were used separately to derive the JC distances (Jukes and Cantor 1969).
Annotated sequences of protein coding genes were also aligned using the Clustal W program. For coding sequences, KS and KA values were calculated using the Li93 method (Li 1993). For noncoding sequences, the JC distances were computed.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionConclusionReferences |
|---|
Alignment of Genomic Sequences
An example of alignment output is shown in Figure 1a. This region contains a FRAM repeat (the underlined part). The nonrepetitive segments before and after the FRAM sequence were used as anchors so that the FRAM in this region of the human genome was not aligned to a FRAM in another part of the chimpanzee genome. Each contig contains hundreds of anchors, greatly facilitating the alignments and increasing the alignment speed. Overall the alignment strategy worked well for all the contigs retrieved. However, there are still two problems in aligning genomic sequences. The first one is the variation in repeat boundaries determined by the RepeatMasker. An example is shown in Figure 1b. The underlined segment of the human sequence was identified as part of an Alu element, whereas its chimpanzee counterpart was treated as a nonrepetitive sequence by the RepeatMasker. Since repeats and nonrepeats are aligned separately in our program, the length variation inevitably leads to gaps in the alignment. It is clear that the gaps in Figure 1b are incorrectly assigned, because the two sequence segments corresponding to the gaps can be easily aligned simply by deleting the gaps. These gaps, however, constitute only a small portion of the total length. Therefore, excluding these gaps from the genetic distance calculations has no significant effect on the genetic distance estimation.
|
The second problem is the occasional occurrence of "abnormal" alignments. An example is shown in Figure 1c. Such alignments are abnormal for three reasons: First, they occur in nonrepetitive sequences, which are usually more stable than repetitive sequences. Second, they have much higher genetic distances than their surrounding sequences. Chimpanzee contig AC087603, for example, shows an increase in the nonrepetitive JC distance, from 1.11% to 7.56%, with only a 1.9 kb (8.8%) increase in total length when two abnormal alignments are included (data not shown). Third, the alignments are always accompanied by multiple gaps. Such abnormal alignments appear in five of the 26 chimpanzee contigs compared: AC087601, AC087603, AC087568, AC087562, and AC087604 (Table 1). The lengths of these abnormal alignments are 2601 bp, 1920 bp, 4110 bp, 3234 bp, and 4775 bp, respectively. It is worth noting that AC087601 and AC087603 belong to the same chromosome (chromosome 3), as do contigs AC087604 and AC087562 (chromosome 12). These two chromosomes comprise only 7% of the total length studied, yet show most of these abnormal alignments. As will be discussed later, these abnormal alignments probably represent the alignments of two nonhomologous sequences and were not included in our analysis.
|
Genomic Sequence Divergence
The genomic sequence divergence between human and chimpanzee is shown in Table 1. This study used 26 relatively long contigs, most of which were longer than 40 kb, for genetic distance calculation. The alignments of 2.30 Mb of human-chimpanzee homologous sequences, including repetitive elements, show an average JC distance of 1.22%. If the repeats are excluded, the total length decreases to 1.44 Mb and the JC distance becomes 1.14%. The overall average and the nonrepetitive average genetic distances are close to Chen and Li's (2001) estimation of 1.24%, which was derived from 53 short (
500 bp), nonrepetitive sequence segments. In comparison, the aligned repetitive elements, with a total length of 0.86 Mb, give a JC distance of 1.34%. The difference between repetitive and nonrepetitive sequence JC distances is significant. Individual chromosomes differ greatly from each other in genetic distance. For example, chromosome 12 shows a distance of 2.43% in repetitive elements, compared with the 1.31% distance for repetitive elements on chromosome 7. For nonrepetitive sequences, the 1.04% distance for chromosome 11 is significantly lower than the 1.42% distance for chromosome 12. Most of the pairwise distance differences between chromosomes are significant. The nonsignificant pairs are the overall and repetitive distances between chromosomes 7 and 11, the nonrepetitive distance between chromosomes 3 and 12, and the repetitive distances between chromosomes 3, 7, and 11.
Similarly the genetic distances for individual contigs vary considerably, even if the contigs belong to the same chromosome. When the short contig AC087602 (chromosome 3) is disregarded, the nonrepetitive distances range from 0.76% (AC087513, chromosome 7) to 1.59% (AC091296, chromosome 7), and the repetitive element distances range from 0.94% (AC087730, chromosome 7) to 3.49% (AC087604, chromosome 12). It is clear that the substitution rates differ considerably from region to region. The variation is especially significant when the sequence is short, as shown in the extreme genetic distances on contig AC087602.
Coding Region Distances
The KS and KA values of 88 coding genes are shown in Table 2. The average KS (1.46%) value is significantly higher than the average genomic JC distance (1.22%), though the average KA value (0.56%) is relatively low (Table 2). The dataset is further divided into three subgroups: single-copy genes, duplicate genes, and immune-related genes. The KS averages of these three subgroups are 1.40%, 1.83%, and 1.32%, respectively, and the KA averages are 0.46%, 0.78%, and 0.59%, respectively. The immune-related gene subgroup has a lower average KS, though a higher average KA, than those of the single-copy genes. Therefore the immune-related genes actually decrease the overall average KS, though they slightly increase the KA average. The duplicate genes, on the other hand, significantly increase the overall KS and KA averages because they have both the highest average KS and KA.
|
It is interesting to note that the KS values fluctuate greatly within each subgroup. In the single-copy gene subgroup, for example, the KS values range from 0 to 4.08%. In the other two subgroups the KS values fluctuate between 0 and 3.50%. Therefore none of the three subgroups is homogeneous in terms of substitution rate.
Figure 2 shows the distribution of KS and KA values for the 88 coding genes. The KS values spread through a wider range of substitution rates and are more evenly distributed in each distance interval than are KA values. More than 65% of the genes have KS values smaller than 2.0%, and thus the KS values are more concentrated in the lower (left) half of the distribution (Figure 2a). The number of genes with KS values between 0.0% and 1.0% is approximately the same as the number of genes with KS values between 1.0% and 2.0% (31 genes and 29 genes, respectively), but the number of genes with KS values between 2.0% and 3.0% is one-third smaller (20 genes) than in the previous two intervals. Only 10 genes have KS values larger than 3.0%. For the KA value distribution, the number of genes decreases sharply as the KA value goes up (Figure 2b). About 60% of the genes have KA values smaller than 0.5%, and more than 80% of the genes have KA values smaller than 1.0%. Nevertheless, there is still a small group of genes (six genes) that have KA values larger than 2.0%. These genes are Ig
1 heavy chain constant region, haptoglobin, transition protein 2, protamine 1, and protamine 2. The last three genes are related to spermatogenesis and are suggested to be under male-driven positive selection (Wyckoff et al. 2000). The haptoglobin gene is a duplicate gene.
|
Divergences in Noncoding Regions of Genes
The JC distances between human and chimpanzee in intron, 5' flanking, 3' flanking, promoter, and pseudogene sequences were also calculated (Table 3). It is important to note that these regions were identified and categorized based on GenBank annotations. In addition, the distances of 5' and 3' flanking regions have to be taken with caution because the boundaries of these regions are not well defined. The "5' flanking region" in this study starts from exactly one base upstream of the start codon and extends to the 5' end of the sequence retrieved. A similar comment applies to the "3' flanking region." Therefore the "flanking regions" may contain 5' and 3' untranslated regions (UTRs) and/or promoters.
|
Among the noncoding regions studied, the 3' flanking region has the highest average JC distance (1.68%), followed in decreasing order by introns (1.47%), 5' flanking regions (1.41%), pseudogenes (1.39%), and promoters (0.75%). The average of all these regions is 1.44%. It is noteworthy that most of the genes in Table 3 also appear in Table 2 and that both the average KS and the average noncoding distances are significantly higher than the average genomic distances in Table 1. The significantly lower value in the promoter regions is expected, given their functional constraint. The 3' flanking regions have on average a higher distance than introns, 5' flanking regions, and pseudogenes, which all are very similar to the KS values observed within coding sequences. However, no definite conclusion can be made because of the relatively high standard errors in the genetic distances for 5', 3', and promoter regions.
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionConclusionReferences |
|---|
Implications for Large-Scale Genomic Studies
This study is the first attempt to incorporate repetitive elements in large-scale genomic alignments. In doing so, several challenges have been met and overcome. First, repetitive elements greatly confound the alignment procedure. Repetitive elements spread throughout the human and chimpanzee genomes and generate considerable amounts of noise in sequence alignments. In this study the length of repetitive elements aligned is 866 kb, approximately 38% of the total length compared. Even though repeats can be excluded before alignment, there still exist some problems, such as uncertainties about the boundaries of repetitive elements, time-consuming repeat masking processes, and erroneous alignments in the vicinity of repeats.
Second, evolutionary events, such as large-scale indels, translocations, and chromosome fusions, can easily lead to fragmented or misleading alignments. Note that indels can often arise because of insertion or deletion of repetitive elements.
Third, alignment programs give erroneous alignments occasionally, and these errors distort distance calculations considerably. For example, under certain situations Clustal W's default penalty settings favor two single-base gaps over one transversion event. In a 100 kb-long alignment, such errors may happen tens of times and lead to an incorrect distance calculation. In the case of large-scale alignments, it is almost impossible to check the alignments manually. Hence, despite the low genetic distance between human and chimpanzee sequences, the alignment task is actually fairly complicated.
The alignment strategy adopted in this study is proven to be effective, though relatively time-consuming. When repeats are excluded from the sequences, the alignment of the resulting sequences may occasionally be incorrect. Because the repeat-deleted segments of a contig are concatenated to form a substitute sequence for alignment, the "welding spots" at the segment joints may lead to false local alignments. To correct this problem, repeats were reinserted back into the aligned repeat-free sequences using the coordinates of the nonrepeat alignments to find homologous repetitive elements. The information from the aligned repeats was then used to correct the nonrepetitive alignment. Most of our alignments were problem-free. Some "abnormal" cases, however, still occurred. For the abnormal alignments mentioned in the Results section, some were located in the vicinity of repetitive elements or within contig gaps. It is likely that the sequences compared are in fact nonhomologous. In addition, as the genetic distance spreads evenly across each contig, it is unlikely to have one or two highly mutable short segments standing amid a sea of conserved sequences. Therefore we concluded that these abnormal alignments were not reliable and they were excluded before calculating genetic distances.
Genetic Distance Between Human and Chimpanzee
The average distance between human and chimpanzee nonrepetitive sequences obtained in the current study is 1.14%, slightly lower than the estimate of 1.24% obtained by Chen and Li (2001). However, the two datasets are different in that Chen and Li's (2001) dataset did not include known genes or predicted genes within the sequences compared, while the dataset in this study does. Therefore the JC distances obtained in this study tend to be smaller because of the potential functional constraints of the coding regions included.
Repetitive elements show slightly higher mutation rates than nonrepetitive sequences, possibly due to a higher proportion of CpG dinucleotides and length variation. For example, Alus are known to have higher CpG contents than other noncoding, nonrepetitive sequences and evolve at a faster pace (Chen and Li 2001). Moreover, repeats have variable lengths. Under certain circumstances the different lengths may be mistaken as mismatches rather than gaps in sequence alignments. In addition, the increased rate of recombination and unequal crossover among tandem repeats predisposes repetitive elements to frequent sequence gains or losses. As a consequence, repeats may tend to have higher genetic distances than do nonrepeats. We note further that the RepeatMasker program sometimes "recognizes" different boundaries for two homologous repetitive elements. Therefore some parts of the two homologous elements are staggered apart in the alignment and erroneous gaps are assigned at the ends, as illustrated in Figure 1b. However, such errors would have little effect on our distance estimation, because gaps were excluded in our analysis.
Another interesting observation is that genetic distances vary considerably among different genomic regions. For example, the JC distances for chromosome 12 are significantly higher than those for the other three chromosomes studied (Table 1). Moreover, the genetic distances fluctuate from region to region, even when they are located on the same chromosome. For example, the chromosome 7 contigs have genetic distances ranging from 0.84% to 1.45% (Table 1). For nonrepetitive distances, the highest (1.59% of AC091296) is more than twice the lowest (0.76% of AC087513). Hence any attempt to derive a representative genetic distance between two species must include multiple regions from different chromosomes. Since most sequences are noncoding, the large variation in substitution rate among regions on the same chromosome or on different chromosomes suggests significant variations in mutation ratios among regions.
For coding sequences, the average KS (1.46%) value obtained in this study is unexpectedly high (Table 2) in comparison with the average genomic divergence. Among the three subgroups, the duplicate gene subgroup has the highest average KS and KA values. For example, the FUT gene cluster that includes FUT3, FUT5, and FUT6 has KS values all larger than 3.00%. These genes were reported to result from successive duplications and the subsequent functional divergence about 10 million years ago (Costache et al. 1997). With a total length of 3324 bp, the FUT3-5-6 cluster increases the overall average KS value by 0.08%.
One possible reason for the high KS values in the duplicate gene subgroup is that some of the genes either are under positive selection or have relaxed purifying selection pressure. A second possible reason is that paralogous genes might have been mistaken as orthologous genes. A third reason is that gene conversion might have happened between duplicate genes. For divergent duplicate genes, gene conversion will considerably increase the genetic distances. Therefore the KS and KA values of duplicate genes should be taken with caution.
The immune-related genes, on the other hand, are an assembly of very heterogeneous genes in terms of substitution rate. Some of them are well conserved, even better conserved than nonimmune genes. For example, complement C5 receptor shows both KS and KA values lower than the 88-gene averages. In addition, the Ig
constant region has KS and KA both equal to zero. In contrast, the Ig1 heavy chain constant region has very high KS and KA values.
For the noncoding regions, the average JC distances in introns, 5' flanking regions, 3' flanking regions, and pseudogenes are all higher than the genomic average of 1.22%. As mentioned in the result section, most of the genes listed in Table 3 also appear in Table 2. The concurrent large divergences in exons, introns, and flanking regions of the same gene may be attributed to high mutation pressure in the regions studied. For example, preproinsulin has a KS value of 4.08% and the corresponding JC distances in the intron, 5' flanking, and 3' flanking regions are 2.23%, 3.75%, and 2.81%, respectively, indicating a higher-than-average mutation pressure. In comparison, glucocerobrosidase shows a KS value of 1.27% and JC distances of 1.12%, 1.00%, and 0.88% in introns, 5' flanking regions, and 3' flanking regions, respectively. The hypothesis of variation in mutation rate among regions may also explain why the average JC distances in the introns and the flanking regions are all higher than that in the pseudogenes, which are expected to diverge more than other genomic regions because of high CpG contents.
|
Conclusion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionConclusionReferences |
|---|
Large-scale alignments of genomic sequences between closely related species, such as human and chimpanzee, can be effectively handled by the strategy employed in this study. For more distant species, however, our strategy may not be effective, because the genomic distribution of repetitive elements may be very different between the two species under study. We are developing a better method to deal with more divergent genomes.
For the human-chimpanzee divergence, the genomewide genetic distances of 1.1% for nonrepeats and 1.2% for the overall average are very close to the estimate (1.24%) of Chen and Li (2001). The fact that two studies using different methods and sequence data give very similar estimates increases our confidence in the estimate. On the other hand, the average KS value of 1.46% from 88 coding genes suggests a higher-than-average mutation pressure on the genes studied. The assumption of high mutation rates is further supported by the higher genetic distances from introns, 5' flanking regions, and 3' flanking regions than those from the genomic sequences. In contrast, the relatively low KA values imply strong purifying selection. The dataset used in this study, though much larger than those of previous studies, is still a small part of the chimpanzee and human genomes. Future studies that include more comprehensive datasets and population polymorphism data can further clarify whether high mutation rates and positive selection play an important role in human evolution or not. Further, the differences in number, type, and influence of repetitive elements should be studied in more detail to reveal the so far obscure, but probably rich information hidden in almost half of the human genome.
|
Acknowledgments |
|---|
This study was supported by NIH grants GM55795, HD38287, and GM30998. F.-C. Chen was supported by the King Car Foundation of Taiwan.
|
Footnotes |
|---|
This paper was delivered at a symposium entitled "Primate Evolutionary Genetics" sponsored by the American Genetic Association at Town and Country Resort and Convention Center, San Diego, CA, USA, May 19–20, 2001.
Corresponding Editor: Oliver A. Ryder
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionConclusionReferences |
|---|
-
Arcot SS, DeAngelis MM, Sherry ST, Adamson AW, Lamerdin JE, Deininger PL, Carrano AV, and Batzer MA, 1997. Identification and characterization of two polymorphic Ya5 Alu repeats. Mutat Res 382:5–11.[Web of Science][Medline]
Bailey WJ, Fitch DH, Tangle DA, Czelusniak J, Slightom JL, and Goodman M, 1991. Molecular evolution of the 
-globin gene locus: gibbon phylogeny and the hominoid slowdown. Mol Biol Evol 8:155–184.[Abstract]
Batzer MA, Stoneking M, Alegria-Hartman M, Bazan H, Kass DH, Shaikh TH, Novick GE, Ioannou PA, Scheer WD, Herrera RJ, et al. 1994. African origin of human-specific polymorphic Alu insertions. Proc Natl Acad Sci USA 91:12288–12292.
Bohossian HB, Skaletsky H, and Page DC, 2000. Unexpectedly similar rates of nucleotide substitution found in male and female hominids. Nature 406:622–625.[Medline]
Chen FC and Li WH, 2001 Genomic divergences between humans and other hominoids and the effective population size of the common ancestor of humans and chimpanzees. Am J Hum Genet 68:444–456.[Web of Science][Medline]
Costache M, Apoil PA, Cailleau A, Elmgren A, Larson G, Henry S, Blancher A, Iordachescu D, Oriol R, and Mollicone R, 1997. Evolution of fucosyltransferase genes in vertebrates. J Biol Chem 272:29721–29728.
Gu Z, Wang H, Nekrutenko A, and Li WH, 2000. Densities, length proportions, and other distributional features of repetitive sequences in the human genome estimated from 430 megabases of genomic sequence. Gene 259:81–88.[Web of Science][Medline]
Horai S, Hayasaka K, Kondo R, Tsugane K, and Takahata N, 1995. Recent African origin of modern humans revealed by complete sequences of hominoid mitochondrial DNAs. Proc Natl Acad Sci USA 92:532–536.
Jukes TH and Cantor CR, 1969. Evolution of protein molecules. In: Mammalian protein metabolism (Munro HN, ed). New York: Academic Press; 21–123.
Kaessmann H, Wiebe V, and Pääbo S, 1999. Extensive nuclear DNA sequence diversity among chimpanzees. Science 286:1159–1162.
Li WH, 1993. Unbiased estimation of the rates of synonymous and nonsynonymous substitution. J Mol Evol 36:96–99.[Web of Science][Medline]
Li WH, Gu Z, Wang H, and Nekrutenko A, 2001. Evolutionary analyses of the human genome. Nature 409:847–849.[Medline]
Nachman MW and Crowell SL, 2000. Estimate of the mutation rate per nucleotide in humans. Genetics 156:297–304.
Rowold DJ and Herrera RJ, 2000. Alu elements and the human genome. Genetica 108:57–72.[Web of Science][Medline]
Sibley CG and Ahlquist JE, 1987. DNA hybridization evidence of hominoid phylogeny: results from an expanded data set. J Mol Evol 26:99–121.[Web of Science][Medline]
Smit AF, 1999. Interspersed repeats and other mementos of transposable elements in mammalian genomes. Curr Opin Genet Dev 9:657–663.[Web of Science][Medline]
Szmulewicz MN, Novick GE, and Herrera RJ, 1998. Effects of Alu insertions on gene function. Electrophoresis 19(8–9):1260–1264.
Thompson JD, Higgins DG, and Gibson TJ, 1994. CLUSTAL W: improving the sensitivity of progressive multiple sequence alignment through sequence weighting, position-specific gap penalties and weight matrix choice. Nucleic Acids Res 22:4673–4680.
Wyckoff GJ, Wang W, and Wu CI, 2000. Rapid evolution of male reproductive genes in the descent of man. Nature 403:304–309.[Medline]
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  S. Kumar, A. Filipski, V. Swarna, A. Walker, and S. B. Hedges Placing confidence limits on the molecular age of the human-chimpanzee divergence PNAS, December 27, 2005; 102(52): 18842 - 18847. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 D. J. Gaffney and P. D. Keightley The scale of mutational variation in the murid genome Genome Res., August 1, 2005; 15(8): 1086 - 1094. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
Y. Gilad, S. A. Rifkin, P. Bertone, M. Gerstein, and K. P. White Multi-species microarrays reveal the effect of sequence divergence on gene expression profiles Genome Res., May 1, 2005; 15(5): 674 - 680. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
Y. Gilad, O. Man, and G. Glusman A comparison of the human and chimpanzee olfactory receptor gene repertoires Genome Res., February 1, 2005; 15(2): 224 - 230. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 X. Wang, S. D. Thomas, and J. Zhang Relaxation of selective constraint and loss of function in the evolution of human bitter taste receptor genes Hum. Mol. Genet., November 1, 2004; 13(21): 2671 - 2678. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. Fischer, V. Wiebe, S. Paabo, and M. Przeworski Evidence for a Complex Demographic History of Chimpanzees Mol. Biol. Evol., May 1, 2004; 21(5): 799 - 808. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. L. Sonnenburg, T. K. Altheide, and A. Varki A uniquely human consequence of domain-specific functional adaptation in a sialic acid-binding receptor Glycobiology, April 1, 2004; 14(4): 339 - 346. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 N. Yu, M. I. Jensen-Seaman, L. Chemnick, J. R. Kidd, A. S. Deinard, O. Ryder, K. K. Kidd, and W.-H. Li Low Nucleotide Diversity in Chimpanzees and Bonobos Genetics, August 1, 2003; 164(4): 1511 - 1518. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
D. E. Wildman, M. Uddin, G. Liu, L. I. Grossman, and M. Goodman Inaugural Article: Implications of natural selection in shaping 99.4% nonsynonymous DNA identity between humans and chimpanzees: Enlarging genus Homo PNAS, June 10, 2003; 100(12): 7181 - 7188. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. Subramanian and S. Kumar Neutral Substitutions Occur at a Faster Rate in Exons Than in Noncoding DNA in Primate Genomes Genome Res., May 1, 2003; 13(5): 838 - 844. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
R. Sakate, N. Osada, M. Hida, S. Sugano, I. Hayasaka, N. Shimohira, S. Yanagi, Y. Suto, K. Hashimoto, and M. Hirai Analysis of 5'-End Sequences of Chimpanzee cDNAs Genome Res., May 1, 2003; 13(5): 1022 - 1026. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
G. Liu, N. C. S. Program, S. Zhao, J. A. Bailey, S. C. Sahinalp, C. Alkan, E. Tuzun, E. D. Green, and E. E. Eichler Analysis of Primate Genomic Variation Reveals a Repeat-Driven Expansion of the Human Genome Genome Res., March 1, 2003; 13(3): 358 - 368. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
R. C. Hardison, K. M. Roskin, S. Yang, M. Diekhans, W. J. Kent, R. Weber, L. Elnitski, J. Li, M. O'Connor, D. Kolbe, et al. Covariation in Frequencies of Substitution, Deletion, Transposition, and Recombination During Eutherian Evolution Genome Res., January 1, 2003; 13(1): 13 - 26. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. Yi, D. L. Ellsworth, and W.-H. Li Slow Molecular Clocks in Old World Monkeys, Apes, and Humans Mol. Biol. Evol., December 1, 2002; 19(12): 2191 - 2198. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
N. G.C. Smith, M. T. Webster, and H. Ellegren Deterministic Mutation Rate Variation in the Human Genome Genome Res., September 1, 2002; 12(9): 1350 - 1356. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||