Comparison of the Complete Protein Sets of Worm and Yeast: Orthology and Divergence

C. elegans Sequencing Consortium, Science 282, 2012 (1998).

A. Goffeau et al., ibid. 274, 546 (1996).

Horvitz H. R., Sulston J. E., Genetics 96, 435 (1980);

Sulston J. E., White J. G., Dev. Biol. 78, 577 (1980);

Kenyon C., Science 240, 1448 (1988);

Sternberg P. W., Adv. Genet. 27, 63 (1990);

Sternberg P. W., Felix M. A., Curr. Opin. Genet. Dev. 7, 543 (1997).

Nasmyth K., Shore D., Science 237, 1162 (1987);

Herskowitz I., Microbiol. Rev. 52, 536 (1988);

Marsh L., Herskowitz I., Cold Spring Harbor Symp. Quant. Biol. 53, 557 (1988) ;

Marsh L., Neiman A. M., Herskowitz I., Annu. Rev. Cell Biol. 7, 699 (1991);

Nasmyth K., Trends Genet. 12, 405 (1996).

J. Gerhart and M. Kirschner, Cells, Embryos, and Evolution (Blackwell, Malden, MA, 1997).

Doolittle R. F., Annu. Rev. Biochem. 64, 287 (1995).

Orthology is not necessarily a one-to-one relationship; a unique gene in one species may be the ortholog of a gene family in another species. The issue is further confounded by the fact that many proteins, particularly in eukaryotes, contain multiple domains that have a degree of evolutionary independence and are found in different combinations. Thus, the detection of even very high similarity between protein sequences from two species does not guarantee that the proteins in question are genuine orthologs with a conserved domain architecture (10). To diminish (but not eliminate) the latter problem, we required that the members of each protein pair be aligned through at least 80% of their lengths.

Fitch W. M., Syst. Zool. 19, 99 (1970).

Koonin E. V., Mushegian A. R., Galperin M. Y., Walker D. R., Mol. Microbiol. 25, 619 (1997).

Hennikof S., et al., Science 278, 609 (1997);

; R. L. Tatusov, E. V. Koonin, D. J. Lipman, ibid., p. 631; M. Y. Galperin and E. V. Koonin, Silico Biol. 1, 7 (1998) (www.bioinfo.de/isb/1998/01/0007).

Doolittle R. F., et al., Science 271, 470 (1996);

Feng D.F., Cho G., Doolittle R. F., Proc. Natl. Acad. Sci. U.S.A. 94, 13028 (1997).

The data set used for these comparisons was the 16 October 1998 worm protein data set from the Sanger Centre and the ORF translations in the 28 October version of the SGD (both are available from the Science Web site as well as SGD). Because the prediction of C. elegans protein sequences had, on 16 October, yet to be corrected by rigorous experimental analysis, our reliance on these predictions may result in the loss of some subset of C. elegans proteins. However, using the subset of yeast proteins for which we had identified no worm homolog, we performed BLAST searches against six frame translations of the entire worm DNA sequence (finished sequence from the Sanger Centre Web site as of 3 November 1998) and identified no additional homologs at the P< 10⁻¹⁰ level with the >80% alignment requirement (J. M. Cherry, unpublished data). Supplemental information regarding the analysis is available at www.sciencemag.org/feature/data/c-elegans.shl for a general overview and at www.sciencemag.org/feature/data/985134.shl for information specific to this review.

Altschul S. F., Gish W., Miller W., Myers E. W., Lipman D. J., J. Mol. Biol. 215, 403 (1990).

Gish W., States D. J., Nature Genet. 3, 266 (1993);

Karlin S., Altschul S. F., Proc. Natl. Acad. Sci. U.S.A. 90, 5873 (1993);

; S. F. Altschul and W. Gish Methods Enzymol. 266, 460 (1996). Version 2.0a19MP-WashU of BLAST was used, with the XNU and SEG filters, BLOSUM62 scoring matrix, with gapping on, and other parameters set to default values.

Thompson J. D., Higgins D. G., Gibson T. J., Nucleic Acids Res. 22, 4673 (1994);

Higgins D. G., Thompson J. D., Gibson T. J., Methods Enzymol. 266, 383 (1996);

. Version 1.74 of CLUSTALW was used with the BLOSUM substitution matrices.

Green P., et al., Science 259, 1711 (1993).

Clarke N. D., Berg J. M., ibid. 282, 2018 (1998).

Kataoka T., et al., Cell 40, 19 (1985).

Chamberlin H. M., Sternberg P. W., Development 120, 2713 (1994);

Church D. L., Guan K. L., Lambie E. J., ibid. 121, 2525 (1995);

Gutch M. J., Flint A. J., Keller J., Tonks N. K., Hengartner M. O., Genes Dev. 12, 571 (1998);

Sundaram M., Yochem J., Han M., Development 122, 2823 (1996);

Yochem J., Sundaram M., Han M., Mol. Cell. Biol. 17, 2716 (1997).

Botstein D., Chervitz S. A., Cherry J. M., Science 277, 1259 (1997);

Botstein D., Fink G. R., ibid. 240, 1439 (1988).

Mori H., Palmer R. E., Sternberg P. W., Mol. Gen. Genet. 245, 781 (1994).

Although there are strong theoretical reasons for preferring the unrooted tree, we show the rooted trees because they are easier to display compactly and more clearly represent the relationships at the tips of the branches, where the assessment of orthology is made. These are, in fact, just representations of unrooted trees with rooting that should be considered arbitrary.

Felsenstein J., Methods Enzymol. 266, 418 (1996).

Rogalski T. M., Riddle D. L., Genetics 118, 61 (1988);

Archambault J., Friesen J. D., Microbiol. Rev. 57, 703 (1993).

Figure 2A shows the CLUSTALW alignment at P < 10⁻²⁰ because at the P < 10⁻¹⁰ the yeast protein Spt5p is included, paired with a presumed worm ortholog even though the similarity of these to RNA polymerases is, upon further study, clearly spurious. This artifact, due to low complexity in the Spt5p amino acid sequence, is avoidable by more aggressive filtering, applying the >80% alignment requirement as well as increasing the stringency; each of these measures exacts a cost in information as well. It illustrates that any alignment result has to be studied for robustness with regard to both stringency and filtering.

Mossi R., Hubscher U., Eur. J. Biochem. 254, 209 (1998).

Tanaka K., Biochem. Biophys. Res. Commun. 247, 537 (1998).

S. A. Chervitz et al., data not shown.

Clark S. W., Meyer D. I., Nature 359, 246 (1992);

; J. Cell Biol.127,129 (1994).

Herman R. K., Cari C. K., Hartman P. S., Genetics 102, 379 (1982).

Nelson R. J., Ziegelhoffer T., Nicolet C., Werner-Washburne M., Craig E. A., Cell 71, 97 (1992).

Normington K., Kohno K., Kozutsumi Y., Gething M. J., Sambrook J., ibid. 57, 1223 (1989).

Heschl M. F. P., Baillie D. L, Comp. Biochem. Physiol. 96, 633 (1990).

Altschul S. F., et al., Nucleic Acids Res. 25, 3389 (1997).

The mitochondrial ribosomal protein orthologs have been missed by the automatic comparison procedure primarily because they contain nonconserved NH₂-terminal import peptides as well as COOH-terminal tails (28).

Okimoto R., Macfarlane J. L., Wolstenholme D. R., J. Mol. Evol. 39, 598 (1994).

The domains were primarily from the SMART database [

Schultz J., Milpetz F., Bork P., Ponting C. P., Proc. Natl. Acad. Sci. U.S.A. 95, 5857 (1998);

], to which several domains were added. We have not attempted to cite the literature for each domain, but refer the reader to the SMART database. See also www.bork.embl-heidelberg.de/Modules_db/special_annotation_page.html

Bork P., Schultz J., Ponting C. P., Trends Biochem Sci. 22, 296 (1997).

To obtain robust counts for each of the domains in the yeast and worm protein sets, we compared representative sequences of each domain to the nonredundant protein database (National Center for Biotechnology Information, National Institutes of Health, Bethesda, MD) using the PSI-BLAST program, and the resulting position-dependent weight matrices (profiles) were saved. The number of search iterations and the cutoff for inclusion of sequences in the profile were adjusted individually for each domain. For most widespread domains, several profiles were constructed to ensure complete coverage. The profiles were then compared separately to the yeast and worm protein databases. Typically, the random expectation value of 0.01 was used as the criterion for domain identification, but the search results were additionally scrutinized for the conservation of patterns typical of the respective domain, to ensure the elimination of any false positives. The profiles for each of the domains are available at the Web site. They can be obtained by FTP and used for PSI-BLAST searches.

Hall T. M., et al., Cell 91, 85 (1996).

Porter J. A., et al., ibid. 86, 21 (1996).

Dahl E., Koseki H., Balling R., Bioessays 19, 755 (1997);

Ryan A. K., Rosenfeld M. G., Genes Dev. 11, 1207 (1997).

Franz G., Loukeris T. G., Dialektaki G., Thompson C. R., Savakis C., Proc. Natl. Acad. Sci. U.S.A. 91, 4746 (1994).

Y. Nagai et al., FEBS Lett, 418, 23 (1997); L. Aravind and E. V. Koonin, J. Mol. Biol., in press.

Cherry J. M., et al., Nature 387, 67 (1997).

We thank J. Hodgkin, R. Horvitz, J. Kimble, and the editors of Science for the invitation to write this paper, D. Lipman for suggesting the collaborations, and K. Anders for helpful discussions. We are especially grateful to R. Durbin (Sanger Centre) and L. Hillier (Genome Sequencing Center) for providing sequence information and for their cooperation. The SGD is supported by a P41 national resources grant HG01315, from the National Human Genome Research Institute at the U.S. NIH. S.A.C. is supported by training grant PHS HG 00044. T.S. is supported by grant DE-FG02-98ER62558 from the Department of the Environment.