A genomic perspective on protein families

Science

Volumn 278, Issue 5338, 1997, Pages 631-637

  Tatusov, Roman L ^a   Koonin, Eugene V ^a   Lipman, David J ^a  

^a NATIONAL INSTITUTES OF HEALTH   (United States)

Author keywords

[No Author keywords available]

Indexed keywords

ARTICLE; FUNGUS; GENOME; GRAM NEGATIVE BACTERIUM; GRAM POSITIVE BACTERIUM; MULTIGENE FAMILY; NONHUMAN; PHYLOGENY; PRIORITY JOURNAL; PROTEIN ANALYSIS; TAXONOMY;

AMINO ACID SEQUENCE; ARCHAEAL PROTEINS; BACTERIA; BACTERIAL PROTEINS; CONSERVED SEQUENCE; EVOLUTION, MOLECULAR; FUNGAL PROTEINS; GENES, ARCHAEAL; GENES, BACTERIAL; GENES, FUNGAL; METHANOCOCCUS; MULTIGENE FAMILY; PHYLOGENY; PROTEINS; SACCHAROMYCES CEREVISIAE; SPECIES SPECIFICITY;

BACTERIA (MICROORGANISMS); FUNGI; NEGIBACTERIA; POSIBACTERIA;

EID: 0030660581     PISSN: 00368075     EISSN: None     Source Type: Journal    
DOI: 10.1126/science.278.5338.631     Document Type: Article

References (67)

1. R. D. Fleischmann et al., Science 269, 496 (1995).
2. C. M. Fraser et al., ibid. 270, 397 (1995); R. Himmelreich et al., Nucleic Acids Res. 24, 4420 (1996); T. Kaneko et al., DNA Res. 3, 109 (1996); F. R. Blattner et al., Science 277, 1453 (1997).
3. C. M. Fraser et al., ibid. 270, 397 (1995); R. Himmelreich et al., Nucleic Acids Res. 24, 4420 (1996); T. Kaneko et al., DNA Res. 3, 109 (1996); F. R. Blattner et al., Science 277, 1453 (1997).
4. C. M. Fraser et al., ibid. 270, 397 (1995); R. Himmelreich et al., Nucleic Acids Res. 24, 4420 (1996); T. Kaneko et al., DNA Res. 3, 109 (1996); F. R. Blattner et al., Science 277, 1453 (1997).
5. C. M. Fraser et al., ibid. 270, 397 (1995); R. Himmelreich et al., Nucleic Acids Res. 24, 4420 (1996); T. Kaneko et al., DNA Res. 3, 109 (1996); F. R. Blattner et al., Science 277, 1453 (1997).
6. C. J. Bult et al., Science 273, 1058 (1996).
7. A. Goffeau et al., ibid. 274, 546 (1996); H. W. Mewes et al., Nature 387, 7 (1997).
8. A. Goffeau et al., ibid. 274, 546 (1996); H. W. Mewes et al., Nature 387, 7 (1997).
9. C. R. Woese, Curr. Biol. 6, 1060 (1996); G. J. Olsen and C. R. Woese, Cell 89, 991 (1997) ; E. V. Koonin, Genome Res. 7, 418 (1997).
10. C. R. Woese, Curr. Biol. 6, 1060 (1996); G. J. Olsen and C. R. Woese, Cell 89, 991 (1997) ; E. V. Koonin, Genome Res. 7, 418 (1997).
11. C. R. Woese, Curr. Biol. 6, 1060 (1996); G. J. Olsen and C. R. Woese, Cell 89, 991 (1997) ; E. V. Koonin, Genome Res. 7, 418 (1997).
12. E. V. Koonin, A. R. Mushegian, K. E. Rudd, Curr. Biol. 6, 404 (1996); E. V. Koonin and A. R. Mushegian, Curr. Opin. Genet. Dev. 6, 757 (1996).
13. E. V. Koonin, A. R. Mushegian, K. E. Rudd, Curr. Biol. 6, 404 (1996); E. V. Koonin and A. R. Mushegian, Curr. Opin. Genet. Dev. 6, 757 (1996).
14. W. M. Fitch, Syst. Zool. 19, 99 (1970). This definition may not embrace all of the complexity of relationships between genes in different genomes. For example, if genes A and B are paralogs encoded in genome 1, and A' and B' are their respective orthologs in genome 2, what is the appropriate description of the relationship between A and B'? They formally are not paralogs, even though a generalized definition might include such cases. Furthermore, one-to-many and many-to-many orthologous relationships evidently exist.
15. W. M. Fitch, Philos. Trans. R. Soc. London Ser. B 349, 93 (1995).
16. R. L. Tatusov et al., Curr. Biol. 6, 279 (1996).
17. E. V. Koonin, A. R. Mushegian, M. Y. Galperin, D. R. Walker, Mol. Microbiol. 25, 619 (1997).
18. The protein sequences were from the original references (1-4), with modifications (for example, tentative correction of frame-shift errors) and additions (previously unreported predicted genes) made for E. coli (E. V. Koonin and R. L. Tatusov, unpublished observations; K. E. Rudd, personal communication), H. influenzae (9), M. genitalium and M. jannaschii (10), and S. cerevisiae (T. J. Wolfsberg and D. Landsman, personal communication). The list of systematic names for all E. coli genes was provided by K. Rudd, and the names for all yeast genes were provided by T. Wolfsberg and D. Landsman; the H. influenzae genes were renamed as previously described (9); the gene names for the other species were from the original publications. The resulting protein database from complete genomes used in all comparisons contained 4283 sequences from E. coli, 1703 sequences from H. influenzae, 468 sequences from M. genitalium, 677 sequences from M. pneumoniae, 3168 sequences from Synechocystis sp., 1736 sequences from M. jannaschii, and 5932 sequences from S. cerevisiae, totaling 17,967 sequences. This sequence set is available on the World Wide Web at http://www.ncbi.nlm.nih.gov/ COG. All pairwise comparisons between these sequences were performed using the BLASTPGP program, which is based on an enhanced version of the BLAST algorithm and includes analysis of local alignments with gaps (26). Predicted coiled coil regions in protein sequences were masked before the comparison using the batch version of the COILS2 program [A. Lupas, Methods Enzymol. 266, 513 (1996); D. R. Walker and E. V. Koonin, ISMB 5, 333 (1997)], and additionally, regions of low complexity were masked using the SEG program with default parameters [J. C. Wootton and S. Federhen, Methods Enzymol. 266, 554 (1996)]. Before the detection of triangles of BeTs, paralogs were identified as those proteins from the same lineage that showed greater similarity to each other than to any protein from another lineage. For the purpose of triangle formation, paralogs were treated as a group. The algorithm further included verification that the BeTs included in a triangle formed a consistent multiple alignment; triangles that did not contain a conserved motif were disregarded.
19. The protein sequences were from the original references (1-4), with modifications (for example, tentative correction of frame-shift errors) and additions (previously unreported predicted genes) made for E. coli (E. V. Koonin and R. L. Tatusov, unpublished observations; K. E. Rudd, personal communication), H. influenzae (9), M. genitalium and M. jannaschii (10), and S. cerevisiae (T. J. Wolfsberg and D. Landsman, personal communication). The list of systematic names for all E. coli genes was provided by K. Rudd, and the names for all yeast genes were provided by T. Wolfsberg and D. Landsman; the H. influenzae genes were renamed as previously described (9); the gene names for the other species were from the original publications. The resulting protein database from complete genomes used in all comparisons contained 4283 sequences from E. coli, 1703 sequences from H. influenzae, 468 sequences from M. genitalium, 677 sequences from M. pneumoniae, 3168 sequences from Synechocystis sp., 1736 sequences from M. jannaschii, and 5932 sequences from S. cerevisiae, totaling 17,967 sequences. This sequence set is available on the World Wide Web at http://www.ncbi.nlm.nih.gov/ COG. All pairwise comparisons between these sequences were performed using the BLASTPGP program, which is based on an enhanced version of the BLAST algorithm and includes analysis of local alignments with gaps (26). Predicted coiled coil regions in protein sequences were masked before the comparison using the batch version of the COILS2 program [A. Lupas, Methods Enzymol. 266, 513 (1996); D. R. Walker and E. V. Koonin, ISMB 5, 333 (1997)], and additionally, regions of low complexity were masked using the SEG program with default parameters [J. C. Wootton and S. Federhen, Methods Enzymol. 266, 554 (1996)]. Before the detection of triangles of BeTs, paralogs were identified as those proteins from the same lineage that showed greater similarity to each other than to any protein from another lineage. For the purpose of triangle formation, paralogs were treated as a group. The algorithm further included verification that the BeTs included in a triangle formed a consistent multiple alignment; triangles that did not contain a conserved motif were disregarded.
20. The protein sequences were from the original references (1-4), with modifications (for example, tentative correction of frame-shift errors) and additions (previously unreported predicted genes) made for E. coli (E. V. Koonin and R. L. Tatusov, unpublished observations; K. E. Rudd, personal communication), H. influenzae (9), M. genitalium and M. jannaschii (10), and S. cerevisiae (T. J. Wolfsberg and D. Landsman, personal communication). The list of systematic names for all E. coli genes was provided by K. Rudd, and the names for all yeast genes were provided by T. Wolfsberg and D. Landsman; the H. influenzae genes were renamed as previously described (9); the gene names for the other species were from the original publications. The resulting protein database from complete genomes used in all comparisons contained 4283 sequences from E. coli, 1703 sequences from H. influenzae, 468 sequences from M. genitalium, 677 sequences from M. pneumoniae, 3168 sequences from Synechocystis sp., 1736 sequences from M. jannaschii, and 5932 sequences from S. cerevisiae, totaling 17,967 sequences. This sequence set is available on the World Wide Web at http://www.ncbi.nlm.nih.gov/ COG. All pairwise comparisons between these sequences were performed using the BLASTPGP program, which is based on an enhanced version of the BLAST algorithm and includes analysis of local alignments with gaps (26). Predicted coiled coil regions in protein sequences were masked before the comparison using the batch version of the COILS2 program [A. Lupas, Methods Enzymol. 266, 513 (1996); D. R. Walker and E. V. Koonin, ISMB 5, 333 (1997)], and additionally, regions of low complexity were masked using the SEG program with default parameters [J. C. Wootton and S. Federhen, Methods Enzymol. 266, 554 (1996)]. Before the detection of triangles of BeTs, paralogs were identified as those proteins from the same lineage that showed greater similarity to each other than to any protein from another lineage. For the purpose of triangle formation, paralogs were treated as a group. The algorithm further included verification that the BeTs included in a triangle formed a consistent multiple alignment; triangles that did not contain a conserved motif were disregarded.
21. Although the exact solution depends on the amino acid composition and size of the particular proteins, under zero approximation, if B (from genome b) is the BeT for A (from genome a), and C (from genome c) is the BeT for B, the probability that C is the BeT for A by chance is close to 1/N, where N is the number of genes in genome c, or ∼0.001.
22. C. R. Woese, Microbiol. Rev. 51, 221 (1987); _, R. Overbeek, G. J. Olsen, J. Bacteriol. 176, 1 (1994); N. R. Pace, Science 276, 734 (1997). A BeT to a given clade was registered if detected in any of the constituent species, for example, in E. coli or H. influenzae for the Gram-negative bacteria.
23. C. R. Woese, Microbiol. Rev. 51, 221 (1987); _, R. Overbeek, G. J. Olsen, J. Bacteriol. 176, 1 (1994); N. R. Pace, Science 276, 734 (1997). A BeT to a given clade was registered if detected in any of the constituent species, for example, in E. coli or H. influenzae for the Gram-negative bacteria.
24. C. R. Woese, Microbiol. Rev. 51, 221 (1987); _, R. Overbeek, G. J. Olsen, J. Bacteriol. 176, 1 (1994); N. R. Pace, Science 276, 734 (1997). A BeT to a given clade was registered if detected in any of the constituent species, for example, in E. coli or H. influenzae for the Gram-negative bacteria.
25. H. Watanabe and J. Otsuka, Comput. Appl. Biosci. 11, 159 (1995); E. V. Koonin, R. L. Tatusov, K. E. Rudd, Methods Enzymol. 266, 295 (1996).
26. H. Watanabe and J. Otsuka, Comput. Appl. Biosci. 11, 159 (1995); E. V. Koonin, R. L. Tatusov, K. E. Rudd, Methods Enzymol. 266, 295 (1996).
27. A schematic visual representation of the search results was used for this analysis [T. L. Madden, R. L. Tatusov, J. Zhang, Methods Enzymol. 266, 131 (1996)].
28. A single-linkage clustering procedure was used with random match probability, P < 0.001, as the cutoff (14).
29. A searchable database of COGs is available at http:// www.ncbi.nlm.nih.gov/COG. Each COG was assigned a unique identification number, which includes a letter for the functional category (19) and a number (see examples in Fig. 1 and Tables 1 and 2).
30. M. Lonetto, M. Gribskov, C. A. Gross, J. Bacteriol. 174, 3843 (1992).
31. The broad functional categories of proteins were as defined previously (9), except that transcription was separated from replication, recombination, and repair. This classification is a modification of the system originally developed for E. coli proteins [M. Riley, Microbiol. Rev. 57, 862 (1993)].
32. A partially similar representation of some of the protein families from complete genomes has been recently published [R. A. Clayton, O. White, K. A. Ketchum, J. C. Venter, Nature 387, 459 (1997)].
33. R. F. Doolittle, D.-F. Feng, S. Tsang, G. Chao, E. Little, Science 271, 470 (1996).
34. R. Himmelreich et al., Nucleic Acids Res. 25, 701 (1997).
35. A. R. Mushegian and E. V. Koonin, Proc. Natl. Acad. Sci. U.S.A. 93, 10268 (1996).
36. E. V. Koonin, A. R. Mushegian, P. Bork, Trends Genet. 12, 334 (1996).
37. J. N. Weiser, M. Shchepetov, S. T. Chong, Infect. Immun. 65, 943 (1997).
38. J. P. Gogarten et al., Proc. Natl. Acad. Sci. U.S.A. 86, 6661 (1989); N. Iwabe et al., ibid., p. 9355; J. P. Gogarten, E. Hilario, L. Olendzewski, in Evolution of Microbial Life, D. McL. Roberts, P. Sharp, G. Alderson, M. Collins, Eds. (Cambridge Univ. Press, Cambridge, 1996), pp. 267-292.
39. J. P. Gogarten et al., Proc. Natl. Acad. Sci. U.S.A. 86, 6661 (1989); N. Iwabe et al., ibid., p. 9355; J. P. Gogarten, E. Hilario, L. Olendzewski, in Evolution of Microbial Life, D. McL. Roberts, P. Sharp, G. Alderson, M. Collins, Eds. (Cambridge Univ. Press, Cambridge, 1996), pp. 267-292.
40. J. P. Gogarten et al., Proc. Natl. Acad. Sci. U.S.A. 86, 6661 (1989); N. Iwabe et al., ibid., p. 9355; J. P. Gogarten, E. Hilario, L. Olendzewski, in Evolution of Microbial Life, D. McL. Roberts, P. Sharp, G. Alderson, M. Collins, Eds. (Cambridge Univ. Press, Cambridge, 1996), pp. 267-292.
41. S. F. Altschul et al., Nucleic Acids Res. 25, 3389 (1997). The probability of a random match, P < 0.001, was used in all PSI-BLAST searches.
42. J. E. Walker, M. Saraste, M. J. Runswick, N. J. Gay, EMBO J. 1, 945 (1982); A. E. Gorbalenya and E. V. Koonin, Nucleic Acids Res. 17, 8413 (1989); M. Saraste, P. R. Sibbald, A. Wittinghofer, Trends Biochem. Sci. 15, 430 (1990).
43. J. E. Walker, M. Saraste, M. J. Runswick, N. J. Gay, EMBO J. 1, 945 (1982); A. E. Gorbalenya and E. V. Koonin, Nucleic Acids Res. 17, 8413 (1989); M. Saraste, P. R. Sibbald, A. Wittinghofer, Trends Biochem. Sci. 15, 430 (1990).
44. J. E. Walker, M. Saraste, M. J. Runswick, N. J. Gay, EMBO J. 1, 945 (1982); A. E. Gorbalenya and E. V. Koonin, Nucleic Acids Res. 17, 8413 (1989); M. Saraste, P. R. Sibbald, A. Wittinghofer, Trends Biochem. Sci. 15, 430 (1990).
45. Protein sequences can be submitted for searching against COGs at http://www.ncbi.nlm.nih.gov/ COG/cognitor.html
46. F. C. Bernstein et al., J. Mol. Biol. 112, 535 (1977).
47. G. Chanfreau, S. M. Noble, C. Guthrie, Science 274, 1511 (1996); A. Jenny, L. Minvielle-Sebastia, P. J. Preker, W. Keller, ibid. 274, 1514 (1996); G. Stumpf and H. Domdey, ibid., p. 1517.
48. G. Chanfreau, S. M. Noble, C. Guthrie, Science 274, 1511 (1996); A. Jenny, L. Minvielle-Sebastia, P. J. Preker, W. Keller, ibid. 274, 1514 (1996); G. Stumpf and H. Domdey, ibid., p. 1517.
49. G. Chanfreau, S. M. Noble, C. Guthrie, Science 274, 1511 (1996); A. Jenny, L. Minvielle-Sebastia, P. J. Preker, W. Keller, ibid. 274, 1514 (1996); G. Stumpf and H. Domdey, ibid., p. 1517.
50. J.-F. Tomb et al., Nature 388, 539 (1997).
51. E. V. Koonin, R. L. Tatusov, M. Y. Galperin, M. N. Rozanov, unpublished observations.
52. P. Green et al., Science 259, 1711 (1993).
53. J. Neuhard and R. A. Kelln, in Escherichia coli and Salmonella: Cellular and Molecular Biology, F. C. Neidhardt et al., Eds. (American Society for Microbiology, Washington, DC, ed. 2, 1996), pp. 580-599.
54. E. C. C. Lin, ibid., pp. 307-342.
55. T. W. Morris, K. E. Reed, J. E. Cronan Jr., J. Bacteriol. 177, 1 (1995).
56. P. Bork, N. P. Brown, H. Hegyi, J. Schultz, Protein Sci. 5, 1421 (1996).
57. D. Richter, E. Niegemann, M. Brendel, Mol. Gen. Genet. 231, 194 (1992); R. Wolter, W. Siede, M. Brendel, ibid. 250, 162 (1996).
58. D. Richter, E. Niegemann, M. Brendel, Mol. Gen. Genet. 231, 194 (1992); R. Wolter, W. Siede, M. Brendel, ibid. 250, 162 (1996).
59. H. Hama, T. Kayahara, W. Ogawa, M. Tsuda, T. Tsuchiya, J. Biochem. 115, 1135 (1994).
60. T. Kobayashi et al., ibid. 98, 101 (1985).
61. A. Colbeau et al., Mol. Microbiol. 8, 15 (1993).
62. M. N. Rozanov and E. V. Koonin, unpublished observations.
63. B. E. Alber and J. G. Ferry, Proc. Natl. Acad. Sci. U.S.A. 91, 6909 (1994); C. Kisker et al., EMBO J. 15, 2323 (1996).
64. B. E. Alber and J. G. Ferry, Proc. Natl. Acad. Sci. U.S.A. 91, 6909 (1994); C. Kisker et al., EMBO J. 15, 2323 (1996).
65. E. V. Koonin, Protein Sci. 4, 1608 (1995); M. N. Rozanov and E. V. Koonin, unpublished observations.
66. E. V. Koonin, Protein Sci. 4, 1608 (1995); M. N. Rozanov and E. V. Koonin, unpublished observations.
67. We thank A. Schaffer for modifying the PSI-BLAST program; R. Walker, H. Watanabe, and M. Rozanov for valuable help with data analysis; K. Rudd, T. Wolfsberg, and D. Landsman for unpublished data; and P. Bork, M. Galperin, M. Gelfand, A. Mushegian, P. Pevzner, M. Roytberg, M. Rozanov, and R. Walker for helpful discussions.