10.1101/gr.5204306 Access the most recent version at doi:
2006 16: 1334-1338; originally published online Oct 25, 2006; Genome Res.
Hervé Tettelin and Martin J. Lercher
Joël Savard, Diethard Tautz, Stephen Richards, George M. Weinstock, Richard A. Gibbs, John H. Werren,
the base of the radiation of Holometabolous insects
Phylogenomic analysis reveals bees and wasps (Hymenoptera) at
"Supplemental Research Data"
This article cites 41 articles, 22 of which can be accessed free at:
Freely available online through the Genome Research Open Access option.
top right corner of the article or
Receive free email alerts when new articles cite this article - sign up in the box at the
go to: Genome ResearchTo subscribe to
© 2006 Cold Spring Harbor Laboratory Press
on November 2, 2006 www.genome.orgDownloaded from
Phylogenomic analysis reveals bees and wasps
(Hymenoptera) at the base of the radiation
of Holometabolous insects
Joël Savard,1Diethard Tautz,1Stephen Richards,2George M. Weinstock,2
Richard A. Gibbs,2John H. Werren,3Hervé Tettelin,4and Martin J. Lercher5,6,7
1Abteilung für Evolutionsgenetik, Institut für Genetik, Universität zu Köln, Köln 50674, Germany;2Human Genome Sequencing
Centre, Baylor College of Medicine, Houston, Texas 77002, USA;3Department of Biology, University of Rochester, New York
14627, USA;4The Institute for Genomic Research, Rockville, Maryland 20850, USA;5Department of Biology and Biochemistry,
University of Bath, Bath BA2 7AY, United Kingdom;6European Molecular Biology Laboratory, 69012 Heidelberg, Germany
Comparative studies require knowledge of the evolutionary relationships between taxa. However, neither
morphological nor paleontological data have been able to unequivocally resolve the major groups of holometabolous
insects so far. Here, we utilize emerging genome projects to assemble and analyze a data set of 185 nuclear genes,
resulting in a fully resolved phylogeny of the major insect model species. Contrary to the most widely accepted
phylogenetic hypothesis, bees and wasps (Hymenoptera) are basal to the other major holometabolous orders, beetles
(Coleoptera), moths (Lepidoptera), and flies (Diptera). We validate our results by meticulous examination of potential
confounding factors. Phylogenomic approaches are thus able to resolve long-standing questions about the phylogeny
[Supplemental material is available online at www.genome.org.]
The four major orders of holometabolous insects (Hymenoptera,
Coleoptera, Lepidoptera, and Diptera) encompass over 45% of all
known animal species (Hammond 1992). While analyses based
on morphological (Kristensen 1999) or individual molecular
markers (such as ribosomal RNA; Whiting 2002b) or mitochon-
drial DNA sequences (Castro and Dowton 2005) have confirmed
the monophyly of these orders, they have been unable to elucidate
most of the interordinal relationships with sufficient confidence.
A close relationship between Diptera (flies) and Lepidoptera
(moths) within the long-recognized Mecopterida assemblage is
generally recovered. However, the affinities of Coleoptera
(beetles) and more particularly of Hymenoptera (wasps and bees)
(Castro and Dowton 2005) remain elusive. In the most widely
accepted phylogenetic hypothesis (Kristensen 1999; Whiting
2002b), a preference is given to a sister-group relationship be-
tween Hymenoptera and Mecopterida, while Coleoptera are
placed at a more basal position as a sister group to the Neurop-
terida, another long-recognized assemblage.
To resolve the phylogenetic relationships of the major ho-
lometabolous orders, we adopt a phylogenomic approach, utiliz-
ing a large number of nuclear genes to maximize phylogenetic
signal over noise (Eisen and Fraser 2003; Rokas et al. 2003; Delsuc
et al. 2005; DeSalle 2005; Philippe et al. 2005a). Such approaches,
based on the simultaneous analysis of a large number of nuclear
genes, have already been shown to be a promising route to un-
derstand deep metazoan relationships (Dopazo and Dopazo
2005; Philippe et al. 2005b). Here, we demonstrate that these
methods are also able to resolve long-standing questions about
the phylogeny of insects.
Using EST sequences to obtain phylogenomic data sets has
proven fruitful, e.g., in the analysis of Eukaryota (Philippe et al.
2004), Amoebae (Bapteste et al. 2002), and Coleoptera relation-
ships (Hughes et al. 2006). The use of EST sequences in phylo-
genomic studies of insects was suggested earlier (Theodorides et
al. 2002), but sufficient data to answer the questions addressed
here has only recently become available.
Our analysis focuses on six holometabolous model species,
for which large scale sequencing projects are available or in
progress. These encompass two dipterans (the fruit fly Drosophila
melanogaster and the mosquito Anopheles gambiae), one lepidop-
teran (the silk moth Bombyx mori), one coleopteran (the flour
beetle Tribolium castaneum), and two hymenopterans (the honey
bee Apis mellifera, and the sibling parasitic wasp species Nasonia
vitripennis and Nasonia giraulti). We further include one orthop-
teran (the grasshopper Locusta migratoria) and one hemipteran
(the pea aphid Acyrthosiphon pisum), both of which are uncon-
tested outgroups to the holometabolous insects based on mor-
phological and molecular markers (Boudreaux 1979; Hennig
1981; Kristensen 1991; Wheeler et al. 2001).
Candidate orthologous clusters were assembled from known or
predicted genes based on a stringent sequence similarity crite-
rion, and were then manually curated to ensure orthology (see
Methods). After removing ambiguously aligned regions, we as-
sembled the remaining sequences into a concatenated alignment
of 33,809 amino acid positions from 185 nuclear genes. As ex-
pected, most genes included here perform housekeeping functions
(see Table S1 of the Supplemental information for a list of genes).
E-mail M.J.Lercher@bath.ac.uk; fax 44-1225-386779.
Article published online before print. Article and publication date are at http://
www.genome.org/cgi/doi/10.1101/gr.5204306. Freely available online
through the Genome Research Open Access option.
16:1334–1338 ©2006 by Cold Spring Harbor Laboratory Press; ISSN 1088-9051/06; www.genome.org
on November 2, 2006 www.genome.orgDownloaded from
This data set supported the topology in Figure 1 regardless of
the phylogenetic methodology (maximum likelihood [Guindon
and Gascuel 2003], Bayesian [Yang and Rannala 1997], or maxi-
mum parsimony [Felsenstein 2004]), with nearly 100% bootstrap
support or 100% posterior probabilities in each case (Table 1).
The previously recognized close relationship of Diptera and Lepi-
doptera is recovered and thus substantiated. However, Hyme-
noptera and not Coleoptera is the most basal of the four major
holometabolous orders. To ensure that our results were not in-
fluenced by an unusually evolving subset of sequences (Gadagkar
et al. 2005), we utilized a bootstrap strategy based on the resam-
pling of genes (Nei et al. 2001), again resulting in strong support
for the topology in Figure 1 (Table 1).
Two common sources of error in phylogenetic reconstruc-
tions are compositional biases (Foster and Hickey 1999) and
long-branch attraction (Felsenstein 1978). Two of our species,
the pea aphid Acyrthosiphon pisum and the honey bee Apis mel-
lifera, have a strongly AT-biased genome. This is reflected in an
overrepresentation of amino acids encoded by AT-rich codons
(Table 2), confirmed by statistical analy-
sis (pairwise ?2-tests, Table 3). The re-
moval of the outlier species Acyrtho-
siphon pisum and Apis mellifera from the
data set resulted in the same well sup-
ported topology (Table 1).
Because long branches in the in-
group are restricted to Diptera and Lepi-
doptera, whose relative positions are un-
contested (Kristensen 1999; Whiting
2002b), long-branch attraction is un-
likely to have influenced the topology.
However, even though the branches of
hymenopterans and Tribolium are short,
it is still conceivable that genes with par-
ticular substitution rates in these species
may have biased the phylogeny. Exclu-
sion of such genes on the basis of a rela-
tive rate test (Tajima 1993) did not
change the tree (Table 1).
Improper outgroup choice can potentially influence the in-
ferred rooting of the ingroup. While our two outgroup species
differ profoundly in evolutionary rate and amino acid composi-
tion, the results remained unchanged when using either Locusta
migratoria or Acyrthosiphon pisum individually as the outgroup
The interordinal relationships among holometabolous insect or-
ders had previously proven to be notoriously difficult to resolve.
However, most researchers assumed a basal split between two
super-orders, the Coleoptera–Neuropterida (including the
beetles) and the Hymenoptera–Mecopterida (including wasps,
flies, and moths) (Kristensen 1999; Whiting 2002b). The tree pre-
sented in Figure 1 necessitates a re-evaluation of this consensual
view. The Mecopterida, encompassing Diptera (flies) and Lepi-
doptera (moths), seem now more closely related to Coleoptera
(beetles) than to Hymenoptera (wasps and bees).
In the present framework, the position of Neuropterida
could not be assessed. Neither previous molecular phylogenies
nor morphological characters allow settlement of this issue;
in particular, wing structure features have been argued to sup-
port a sister-group relationship of Neuropterida with either Co-
leoptera (Hornschemeyer 2002) or with Mecopterida (Kukalová-
Peck and Lawrence 2004). The morphological characters used to
support the traditional Holometabola phylogeny should cer-
tainly be reanalyzed in the light of the relationships presented
Why was the basal position of hymenopterans not discov-
ered in previous molecular phylogenetic studies? A plausible ex-
planation is the lack of resolution power of single molecules
when radiations are old or compressed in time (Rokas et al.
2005). Because the phylogenetic split in question occurred at
least 275 million years ago (Mya) (Ponomarenko 2002; Rasnitsyn
2002), analyses based on a single molecule (e.g., 18S rRNA) did
not provide sufficient resolution (Whiting 2002a). While 60% of
the 185 protein alignments analyzed here were better explained
by the tree in Figure 1 than by the previously assumed tree (based
on likelihood comparisons), only two proteins supported the
basal position of Hymenoptera with a bootstrap support >50%
when analyzed individually (removing these proteins does not
monophyly of Coleoptera (beetles), Lepidoptera (moths), and Diptera
(flies), to the exclusion of Hymenoptera (bees and wasps). Branch lengths
are from maximum likelihood. Numbers report maximum likelihood
bootstrap support (in percent); Bayesian posterior probabilities and maxi-
mum parsimony bootstrap support are ?99% for each branch (Table 1).
Holometabolous phylogenetic relationships, showing the
Support values for the nodes in Figure 1
Maximum likelihood bootstrapa
Relative rate testc
L. migratoria excluded
A. pisum excluded
A. pisum and A. mellifera excluded
Single gene consensusd
Bayesian posterior probabilitiesb
Maximum parsimony bootstrapa
aBootstrap fractions estimated from 1000 maximum likelihood or maximum parsimony calculations (in
bPosterior probabilities estimated from all sampled trees in four independent Markov chain Monte
Carlo calculations (in percent).
cAfter removal of 71 genes that showed significant rate heterogeneity (P < 0.2) between Nasonia and
Tribolium castaneum, and excluding Apis mellifera from the alignment.
dMajority-rule consensus based on 100 bootstrap replicates per gene; very similar results are obtained
when only the best tree contributes for each gene.
Hymenopterans are basal holometabolous insects
on November 2, 2006 www.genome.orgDownloaded from
influence the tree topology, data not shown). Accordingly, a con-
sensus tree based on single gene tree reconstructions yielded the
same topology as in Figure 1, but without strong bootstrap sup-
port (Table 1). The present analysis thus supports the notion that
concatenated sequence trees provide more resolution than con-
sensus gene trees (Rokas et al. 2003; Gadagkar et al. 2005): Com-
bined analysis of a large number of sequences was necessary to
resolve the deep evolutionary relationships among holometabo-
Previous studies based on the simultaneous analysis of
many proteins also failed to recover the topology in Figure 1.
Philippe et al. (2004) analyzed 129 proteins in a maximum like-
lihood framework, resulting in a strongly supported sister-group
relationship between Apis mellifera and Bombyx mori, with Dro-
sophila melanogaster located more basal. This topology is incon-
sistent with the monophyly of Mecopterida, which is well sup-
ported by morphological evidence (Kristensen 1999). It is most
likely a long-branch attraction artifact, brought about by the in-
sufficient number of holometabolous species and the use of a
very divergent lineage (a tick) as the outgroup (Philippe and Laur-
ent 1998). This study highlights the importance of selecting close
outgroup species, and the necessity to test the influence of an
alternative outgroup choice. A second study (Philippe et al.
2005b) recently analyzed 146 proteins, failing to resolve the rela-
tive positions of Coleoptera and Hymenoptera. Here, the poten-
tially large amount of missing data between the insect species
considered might be responsible for the lack of resolution.
The interordinal relationships presented in Figure 1 are in
fact also suggested by examination of the fossil record (Rohden-
dorf and Rasnitsyn 1980), and are supported by phylogenetic
analysis of intron positions (Krauss et al. 2005) as well as wing
characters (Ross 1965; Kukalová-Peck and Lawrence 2004). Fur-
thermore, we note that phylogenies based on 18S rRNA se-
quences also yielded a basal Hymenoptera among holometabo-
lous insects, although not with a credible level of support (Whit-
ing 2002b). The sum of evidence hence supports the present
phylogenomic analysis as a reliable foundation for comparative
analyses of the insect model organisms.
Drosophila melanogaster (Adams et al. 2000), Anopheles gambiae
(Holt et al. 2002), and Apis mellifera peptides were obtained from
Ensembl (www.ensembl.org). Bombyx mori (Mita et al. 2003), Lo-
custa migratoria (Kang et al. 2004), Tribolium castaneum, and Acyr-
thosiphon pisum mRNA sequences were downloaded from NCBI
(ftp.ncbi.nlm.nih.gov). Nasonia vitripennis and Nasonia giraulti
EST data were generated by authors J.H.W. and H.T. All nucleo-
tide data sets were cleaned of vector, mitochondrial and bacterial
contaminations using SeqClean (available from www.tigr.org/
tdb/tgi/software/) before being assembled into nonredundant
contigs with cap3 using default settings (Huang and Madan
1999). All nucleotide data sets were then searched against all
Drosophila melanogaster proteins using BLASTx. The reading
frame from the best hit was assumed to be the correct reading
frame. We then chose the longest run of peptides uninterrupted
by a stop codon as the peptide corresponding to each nucleotide
Identification of orthologs
We performed BLASTp searches of all proteome pairs. Orthologs
were selected based on reciprocal best BLAST hits (Tatusov et al.
1997) using an E-value cut-off of 10?25. A group of sequences
with exactly one member in each species (including either one or
both Nasonia species) was accepted as a candidate orthologous
family if each sequence had each of the other family sequences as
the best BLASTp hit in the respective proteome. This requirement
of all-against-all reciprocal best hits is very stringent, and thus
gives good confidence in the inferred orthology. Multiple se-
quence alignments were performed with MUSCLE (Edgar 2004)
using default settings. Resulting alignments were then manually
curated to ensure completeness and consistency. Poorly con-
served families or clusters potentially containing paralogous se-
quences were discarded. For the sibling Nasonia species, when
orthologous sequences were available for both species, the longer
one was chosen. Alignments were then purged from unreliably
aligned positions as well as gaps with Gblocks (Castresana 2000)
using highly stringent settings, where all sequences flanking a
acids encoded by AT-rich and by GC-rich codons, as well as their
Compositional bias, showing the percentage of amino
Species % FYMINKa
Nasonia vitripennis and
aAT-rich amino acids: F (Phe), Y (Tyr), M (Met), I (Ile), N (Asn), and K (Lys).
bGC-rich amino acids: G (Gly), A (Ala), R (Arg), and P (Pro).
significant compositional bias compared to most other species
?2(upper triangle) and Bonferroni-corrected P-values (lower triangle) showing that Acyrthosiphon pisum and Apis mellifera have a
and N. giraulti
N. vitripennis and N. giraulti
Savard et al.
on November 2, 2006 www.genome.org Downloaded from
block had to be conserved and where blocks smaller than 20
amino acids were discarded. We concatenated the final set of 185
nuclear sequences for phylogenetic analysis, resulting in an
eight-species alignment of 33,809 amino acid positions. The list
of genes included in our analysis is available as Supplemental-
We first analyzed the data in a maximum likelihood framework,
using phyML (Guindon and Gascuel 2003) under an empirical
model of amino acid substitutions (Jones et al. 1992), allowing
for substitution rate variation among sites with a gamma distri-
bution (four rate categories). Branch support values in Table 1 are
from analysis of 1000 bootstrap replicates. Using alternative
models of amino acid evolution (WAG, Whelan and Goldman
2001; and VT, Bapteste et al. 2002) led to the same well supported
Additionally, we estimated the tree in a Bayesian framework,
using MrBayes (Huelsenbeck and Ronquist 2001) and employing
the same model of sequence evolution as above. We ran four
independent searches, each starting from a random tree and sam-
pling every tenth tree over 100,000 generations. Each run had
equilibrated after less than 1000 generations; thus, the first 100
trees were disregarded as burn-in. The independent runs consis-
tently resulted in the same topology; posterior probabilities
(Table 1) were calculated from all sampled trees across indepen-
dent runs. Finally, we also analyzed 1000 bootstrap replicates
under a maximum parsimony criterion, using PROTPARS from
the PHYLIP package (Felsenstein 2004).
Gene bootstrap resampling
To test if the obtained tree was dominated by one or a few dis-
parate genes, we performed maximum likelihood analyses of
1000 bootstrap data sets obtained from the resampling of com-
plete genes (Nei et al. 2001). For each replicate, we drew 185 gene
alignments from the full curated data set described above. Be-
cause genes were not removed from the pool after being chosen,
each bootstrap data set contained some gene alignments more
than once, while others were missing altogether; this is analo-
gous to the widely used bootstrap strategy based on individual
amino acid sites. For each replicate, alignments were then con-
catenated and analyzed with phyML as above.
Relative rate tests
To determine if the relative position of Coleoptera and Hyme-
noptera was caused by rate variation among these orders, we
performed a relative rate test (Tajima 1993) between Nasonia and
Tribolium castaneum. We restricted this analysis and the follow-
ing tree reconstruction to sites that were identical between Lo-
custa migratoria and Acyrthosiphon pisum, and where the ancestral
state could thus be inferred reliably. To be conservative, we re-
moved all genes for which ?2> 1.64, i.e., P < 0.2 (1 degree of
freedom) (Kumar and Hedges 1998). Excluding Apis mellifera, we
concatenated the remaining 114 genes into an alignment of
16,495 amino acid positions. Maximum likelihood bootstrap
analysis was performed as above.
Phylogenetic analysis of individual proteins
We also analyzed each individual protein alignment with the
maximum likelihood method as described above. To analyze
each protein’s support for the tree in Figure 1 compared to the
previously assumed tree (with the positions of hymenopterans
and Tribolium exchanged), we compared the likelihoods calcu-
lated under each topology; statistical support was estimated us-
ing the method of Kishino and Hasegawa (1989) as implemented
in the PAML package (Yang 1997).
Compositional heterogeneity among species pairs was as-
sessed with a ?2-test, where ?mn
fmithe total number of amino acids of type i in the concatenated
sequence for species m. The values in Table 3 are based on those
amino acids that are biased in the GC content of their codons
(FYMINK/GARP, see also Table 2; 9 degrees of freedom). Quali-
tatively very similar results are obtained when using all amino
acids (data not shown).
All phylogenomic analyses and tests were implemented in
Perl scripts, which are available in the Supplemental material.
2= ∑i(fmi? fni)2/(fmi+ fni), with
J.H.W. thanks Wayne Hunter (USDA, ARS) and Phat Dang
(USDA, ARS) for construction of the Nasonia EST libraries. We
wish to thank Shannon K. McWeeney and Csaba Pal for helpful
discussions. J.S. and D.T. acknowledge support through grants
from the HFSPO and the DFG. M.J.L. acknowledges financial sup-
port from the Royal Society and the DFG. The Nasonia EST
project of J.H.W. and H.T. was supported by the 21st Century
Research & Technology Fund.
Adams, M.D., Celniker, S.E., Holt, R.A., Evans, C.A., Gocayne, J.D.,
Amanatides, P.G., Scherer, S.E., Li, P.W., Hoskins, R.A., Galle, R.F., et
al. 2000. The genome sequence of Drosophila melanogaster. Science
Bapteste, E., Brinkmann, H., Lee, J.A., Moore, D.V., Sensen, C.W.,
Gordon, P., Durufle, L., Gaasterland, T., Lopez, P., Muller, M., et al.
2002. The analysis of 100 genes supports the grouping of three
highly divergent amoebae: Dictyostelium, Entamoeba, and
Mastigamoeba. Proc. Natl. Acad. Sci. 99: 1414–1419.
Boudreaux, H.B. 1979. Arthropod phylogeny with special reference to insects.
Wiley, New York.
Castresana, J. 2000. Selection of conserved blocks from multiple
alignments for their use in phylogenetic analysis. Mol. Biol. Evol.
Castro, L.R. and Dowton, M. 2005. The position of the Hymenoptera
within the Holometabola as inferred from the mitochondrial
genome of Perga condei (Hymenoptera: Symphyta: Pergidae). Mol.
Phylogenet. Evol. 34: 469–479.
Delsuc, F., Brinkmann, H., and Philippe, H. 2005. Phylogenomics and
the reconstruction of the tree of life. Nat. Rev. Genet. 6: 361–375.
DeSalle, R. 2005. Animal phylogenomics: Multiple interspecific genome
comparisons. Methods Enzymol. 395: 104–133.
Dopazo, H. and Dopazo, J. 2005. Genome-scale evidence of the
nematode-arthropod clade. Genome Biol. 6: R41.
Edgar, R.C. 2004. MUSCLE: Multiple sequence alignment with high
accuracy and high throughput. Nucleic Acids Res. 32: 1792–1797.
Eisen, J.A. and Fraser, C.M. 2003. Phylogenomics: Intersection of
evolution and genomics. Science 300: 1706–1707.
Felsenstein, J. 1978. Cases in which parsimony or compatibility
methods will be positively misleading. Syst. Zool. 27: 401–410.
Felsenstein, J. 2004. PHYLIP (Phylogeny Inference Package) version 3.6.
Department of Genome Sciences, University of Washington, Seattle,
Foster, P.G. and Hickey, D.A. 1999. Compositional bias may affect both
DNA-based and protein-based phylogenetic reconstructions. J. Mol.
Evol. 48: 284–290.
Gadagkar, S.R., Rosenberg, M.S., and Kumar, S. 2005. Inferring species
phylogenies from multiple genes: Concatenated sequence tree versus
consensus gene tree. J. Exp. Zool. (B Mol. Dev. Evol.) 304: 64–74.
Guindon, S. and Gascuel, O. 2003. A simple, fast, and accurate
algorithm to estimate large phylogenies by maximum likelihood.
Syst. Biol. 52: 696–704.
Hammond, P. 1992. Species inventory. In Global biodiversity (ed. B.
Groombridge), pp. 17–39. Chapman & Hall, London.
Hennig, W. 1981. Insect phylogeny. Wiley, New York.
Holt, R.A., Subramanian, G.M., Halpern, A., Sutton, G.G., Charlab, R.,
Nusskern, D.R., Wincker, P., Clark, A.G., Ribeiro, J.M., Wides, R., et
al. 2002. The genome sequence of the malaria mosquito Anopheles
Hymenopterans are basal holometabolous insects
on November 2, 2006 www.genome.org Downloaded from
gambiae. Science 298: 129–149. Download full-text
Hornschemeyer, T. 2002. Phylogenetic significance of the wing-base of
the Holometabola (Insecta). Zool. Scr. 31: 17–29.
Huang, X. and Madan, A. 1999. CAP3: A DNA sequence assembly
program. Genome Res. 9: 868–877.
Huelsenbeck, J.P. and Ronquist, F. 2001. MRBAYES: Bayesian inference
of phylogenetic trees. Bioinformatics 17: 754–755.
Hughes, J., Longhorn, S.J., Papadopoulou, A., Theodorides, K., de Riva,
A., Mejia-Chang, M., Foster, P.G., and Vogler, A.P. 2006. Dense
taxonomic EST sampling and its applications for molecular
systematics of the Coleoptera (beetles). Mol. Biol. Evol. 23: 268–278.
Jones, D.T., Taylor, W.R., and Thornton, J.M. 1992. The rapid
generation of mutation data matrices from protein sequences.
Comput. Appl. Biosci. 8: 275–282.
Kang, L., Chen, X., Zhou, Y., Liu, B., Zheng, W., Li, R., Wang, J., and
Yu, J. 2004. The analysis of large-scale gene expression correlated to
the phase changes of the migratory locust. Proc. Natl. Acad. Sci.
Kishino, H. and Hasegawa, M. 1989. Evaluation of the maximum
likelihood estimate of the evolutionary tree topologies from DNA
sequence data, and the branching order in hominoidea. J. Mol. Evol.
Krauss, V., Pecyna, M., Kurz, K., and Sass, H. 2005. Phylogenetic
mapping of intron positions: A case study of translation initiation
factor eIF2?. Mol. Biol. Evol. 22: 74–84.
Kristensen, N.P. 1991. The phylogeny of hexapods. In The insects of
Australia: A Textbook for students and research workers (eds. I.D.
Naumann et al.), pp. 125–140. Melbourne University Press, Carlton,
Kristensen, N.P. 1999. Phylogeny of endopterygote insects, the most
successful lineage of living organisms. Eur. J. Entomol. 96: 237–253.
Kukalová-Peck, J. and Lawrence, J.F. 2004. Relationships among
coleopteran suborders and major endoneopteran lineages: Evidence
from hind wing characters. Eur. J. Entomol. 101: 95–144.
Kumar, S. and Hedges, S.B. 1998. A molecular timescale for vertebrate
evolution. Nature 392: 917–920.
Mita, K., Morimyo, M., Okano, K., Koike, Y., Nohata, J., Kawasaki, H.,
Kadono-Okuda, K., Yamamoto, K., Suzuki, M.G., Shimada, T., et al.
2003. The construction of an EST database for Bombyx mori and its
application. Proc. Natl. Acad. Sci. 100: 14121–14126.
Nei, M., Xu, P., and Glazko, G. 2001. Estimation of divergence times
from multiprotein sequences for a few mammalian species and
several distantly related organisms. Proc. Natl. Acad. Sci.
Philippe, H. and Laurent, J. 1998. How good are deep phylogenetic
trees? Curr. Opin. Genet. Dev. 8: 616–623.
Philippe, H., Snell, E.A., Bapteste, E., Lopez, P., Holland, P.W., and
Casane, D. 2004. Phylogenomics of eukaryotes: Impact of missing
data on large alignments. Mol. Biol. Evol. 21: 1740–1752.
Philippe, H., Delsuc, F., Brinkmann, H., and Lartillot, N. 2005a.
Phylogenomics. Annu. Rev. Ecol. Evol. Syst. 36: 541–562.
Philippe, H., Lartillot, N., and Brinkmann, H. 2005b. Multigene analyses
of bilaterian animals corroborate the monophyly of Ecdysozoa,
Lophotrochozoa, and Protostomia. Mol. Biol. Evol. 22: 1246–1253.
Ponomarenko, A.G. 2002. Superorder Scarabaeidea Laicharting, 1781.
Order Coleoptera Linné, 1758. The Beetles. In History of insects (eds.
A.P. Rasnitsyn and D.L.J. Quicke), pp. 164–176. Kluwer Academic
Publishers, Dordrecht, The Netherlands.
Rasnitsyn, A.P. 2002. Cohors Scarabaeiformes Laicharting, 1781. The
Holometabolans. In History of insects (eds. A.P. Rasnitsyn and D.L.J.
Quicke), pp. 157–159. Kluwer Academic Publishers, Dordrecht, The
Rohdendorf, B.B. and Rasnitsyn, A.P. 1980. Historical development of the
class insecta. Nauka Press, Moscow.
Rokas, A., Williams, B.L., King, N., and Carroll, S.B. 2003. Genome-scale
approaches to resolving incongruence in molecular phylogenies.
Nature 425: 798–804.
Rokas, A., Kruger, D., and Carroll, S.B. 2005. Animal evolution and the
molecular signature of radiations compressed in time. Science
Ross, H.H. 1965. A textbook of entomology. Wiley, New York.
Tajima, F. 1993. Simple methods for testing the molecular evolutionary
clock hypothesis. Genetics 135: 599–607.
Tatusov, R.L., Koonin, E.V., and Lipman, D.J. 1997. A genomic
perspective on protein families. Science 278: 631–637.
Theodorides, K., De Riva, A., Gomez-Zurita, J., Foster, P.G., and Vogler,
A.P. 2002. Comparison of EST libraries from seven beetle species:
Towards a framework for phylogenomics of the Coleoptera. Insect
Mol. Biol. 11: 467–475.
Wheeler, W.C., Whiting, M., Wheeler, Q.D., and Carpenter, J.M. 2001.
The phylogeny of the extant hexapod orders. Cladistics 17: 113–169.
Whelan, S. and Goldman, N. 2001. A general empirical model of protein
evolution derived from multiple protein families using a
maximum-likelihood approach. Mol. Biol. Evol. 18: 691–699.
Whiting, M.F. 2002a. Phylogeny of the holometabolous insect orders
based on 18S ribosomal data: When bad things happen to good
data. In Molecular systematics and evolution:Theory and practice (ed.
G.G.R. DeSalle, and W.C. Wheeler), pp. 69–83. Birkhäuser Verlag
AG, Basel, Switzerland.
Whiting, M.F. 2002b. Phylogeny of the holometabolous insect orders:
Molecular evidence. Zool. Scr. 31: 3–15.
Yang, Z. 1997. PAML: A program package for phylogenetic analysis by
maximum likelihood. Comput. Appl. Biosci. 13: 555–556.
Yang, Z.H. and Rannala, B. 1997. Bayesian phylogenetic inference using
DNA sequences: A Markov Chain Monte Carlo method. Mol. Biol.
Evol. 14: 717–724.
Received November 3, 2005; accepted in revised form April 3, 2006.
Savard et al.
1338 Genome Research
on November 2, 2006 www.genome.orgDownloaded from