Predicting the Evolution of Human Influenza A

Fitch W. M., Bush R. M., Bender C. A., Cox N. J., Proc. Natl. Acad. Sci. U.S.A. 94, 7712 (1997).

Bush R. M., Fitch W. M., Bender C. A., Cox N. J., Mol. Biol. Evol. 16, 1457 (1999).

The technique for identifying those codons under positive selection excludes amino acid replacements that may have occurred during culture in eggs in the laboratory by ignoring replacements that occur on the branches joining isolate sequences to the tree.

Sequences were generated at the Centers for Disease Control and Prevention by the protocols described in (1). They have been deposited in GenBank (accession nos. to and to ).

D. L. Swofford, PAUP* (Phylogenetic Analysis Using Parsimony and other methods) version 4.0.0d60.

We used trees constructed from the 1983–1997 data both to derive the set of positively selected codons (2) and to score the prediction tests. This does not bias our scoring because the location of the trunk is not a factor in determining which codons are under positive selection.

Trunk nodes present in at least 50% of the bootstrap replicates were consistently present in nucleotide parsimony and neighbor-joining trees and also in trees constructed from amino acids using either parsimony, neighbor-joining (5), or the protML maximum likelihood routine of Molphy 2.2 (made available by J. Adachi and M. Hasegawa). The method we used for evaluating the success of a prediction test is thus robust with regard to the algorithm and type of data used. Computational restraints prevented applying maximum-likelihood algorithms to our entire data set at once, so we used multiple overlapping subsets of sequences to perform these analyses. The data set was also too large for standard bootstrap analysis, so we estimated bootstrap values by using 10,000 PAUP FastStep bootstrap replicates. This number was determined by increasing the number of replicates until we obtained a consistent set of bootstrap values.

Wilson I. A., Cox N. J., Annu. Rev. Immunol. 8, 737 (1990);

Wiley D. C., Wilson I. A., Skehel J. J., Nature 289, 373 (1981).

Supplementary tables showing the codons used in alternative prediction tests are available at www.sciencemag.org/feature/data/1044188.shl

C. A. Bender and N. J. Cox, unpublished data.

Cox N. J., Bender C. A., Semin. Virol. 6, 359 (1995).

W. Weis et al., Nature 333, 426 (1988).

Supported in part by funds provided by the University of California for the conduct of discretionary research by Los Alamos National Laboratory, conducted under the auspices of the US Department of Energy. We gratefully acknowledge the technical expertise of H. Jing and the critical comments of S. Frank.