The use of Open Reading frame ESTs (ORESTES) for analysis of the honey bee transcriptome
- Francis MF Nunes1, 2,
- Valeria Valente3,
- Josane F Sousa3,
- Marco AV Cunha2,
- Daniel G Pinheiro1, 2,
- Rafaela M Maia3,
- Daniela D Araujo3,
- Maria CR Costa2,
- Waleska K Martins4,
- Alex F Carvalho4,
- Nadia Monesi5,
- Adriana M Nascimento6,
- Pablo MV Peixoto7,
- Maria FR Silva7,
- Ricardo GP Ramos3,
- Luis FL Reis4,
- Emmanuel Dias-Neto8,
- Sandro J Souza4,
- Andrew JG Simpson9,
- Marco A Zago2, 10,
- Ademilson EE Soares1,
- Marcia MG Bitondi6,
- Enilza M Espreafico3,
- Foued S Espindola7,
- Maria L Paco-Larson3,
- Zila LP Simoes6,
- Klaus Hartfelder3, 6Email author and
- Wilson A SilvaJr1, 2
© Nunes et al; licensee BioMed Central Ltd. 2004
Received: 07 June 2004
Accepted: 03 November 2004
Published: 03 November 2004
The ongoing efforts to sequence the honey bee genome require additional initiatives to define its transcriptome. Towards this end, we employed the Open Reading frame ESTs (ORESTES) strategy to generate profiles for the life cycle of Apis mellifera workers.
Of the 5,021 ORESTES, 35.2% matched with previously deposited Apis ESTs. The analysis of the remaining sequences defined a set of putative orthologs whose majority had their best-match hits with Anopheles and Drosophila genes. CAP3 assembly of the Apis ORESTES with the already existing 15,500 Apis ESTs generated 3,408 contigs. BLASTX comparison of these contigs with protein sets of organisms representing distinct phylogenetic clades revealed a total of 1,629 contigs that Apis mellifera shares with different taxa. Most (41%) represent genes that are in common to all taxa, another 21% are shared between metazoans (Bilateria), and 16% are shared only within the Insecta clade. A set of 23 putative genes presented a best match with human genes, many of which encode factors related to cell signaling/signal transduction. 1,779 contigs (52%) did not match any known sequence. Applying a correction factor deduced from a parallel analysis performed with Drosophila melanogaster ORESTES, we estimate that approximately half of these no-match ESTs contigs (22%) should represent Apis-specific genes.
The versatile and cost-efficient ORESTES approach produced minilibraries for honey bee life cycle stages. Such information on central gene regions contributes to genome annotation and also lends itself to cross-transcriptome comparisons to reveal evolutionary trends in insect genomes.
The honey bee, Apis mellifera, occupies a prominent place in biological research due to its social behavior, learning capabilities, haplodiploid mechanism of sex determination, and plasticity in phenotype (caste) and longevity. Thus, it is a model organism for classical and sociogenetic studies. In addition, bees drive a large-scale apicultural industry, and also generate important income in small-scale subsistence beekeeping. And finally, bees are of great economic and ecological relevance for their role as generalist pollinators.
The decision to include the honey bee amongst the current organisms for complete genome sequencing, http://www.hgsc.bcm.tmc.edu/projects/honeybee/ was, therefore, well founded, yet information on its transcriptome is still meager. When starting this study, little over 250 genes were annotated as partial or full length coding sequences, and only about 15,500 expressed sequence tags (mainly 5'-ESTs generated from a normalized bee brain cDNA library ) were available in public databases. Thus, even after sequencing of the honey bee genome will be completed a considerable transcriptome sequencing effort will still be required for unequivocal genome annotation, gene identification, and subsequent functional studies.
We used the ORESTES (Open Reading frame Expressed Sequence Tags) strategy to generate ESTs from different life cycle stages of the honey bee, such as appropriate for a genome annotation initiative. This strategy preferentially generates ESTs of the central, and thus most informative portion of the transcript , and frequently also identifies less abundant mRNAs . The efficacy of the Open Reading frame ESTs strategy, in the context of an organism for which there is limited genomic information, has recently been demonstrated for Schistosoma mansoni .
This cost-efficient approach increased the already existent Apis EST database by 30% new reads. Of the 5,021 ORESTES, only 35.2% matched with previously deposited Apis ESTs. When assembled with the existent Apis ESTs in the NCBI database, the ORESTES sequences extended 66% of the mixed contigs. Together these data indicate that the ORESTES methodology could effectively complement the current efforts towards the definition of the Apis transcriptome.
Results and discussion
Honey bee Open Reading frame ESTs
Apis mellifera Open Reading frame ESTs.
Number of reads
Total analyzed reads
- Larval stages
- Stage mix
Local alignment matches
- Apis ESTs from dbEST
- Apis mellifera sequences in GenBank
- Genes of other organisms (orthologs)
- No matches in GenBank
- Number of contigs
- Number of singlets
- Total number of clusters
Gene Ontology classification
Gene Ontology classification of Apis mellifera ORESTES contigs according to the Drosophila genes that they represent.
Number of genes
response to stress
cell growth and/or maintenance
response to external stimulus
nucleic acid binding
RNA polymerase II transcription factor activity
antimicrobial peptide activity
receptor signaling protein activity
structural constituent of cytoskeleton
microfilament motor activity
transcription factor activity
metal ion binding
ion transporter activity
Clustering of honey bee ESTs
Within the total contig population, 9.5% of the assembled sequences (323) are represented by mixed contigs of both AmORESTES and AmNCBI sequences, and within this group 66.3% (214) of the original AmNCBI contigs were considerably extended, or were joined across gaps by the AmORESTES contigs, as illustrated in Figure 2C. The fact that the number of mixed contigs is relatively low compared to total contig number may be attributed to two aspects. First, most of the AmNCBI contigs were obtained from a single tissue (brain) library, whereas the AmORESTES sequences represent whole body transcripts of all life cycle stages of the honey bee. Second, the AmNCBI sequences are mainly 5'-ESTs, whereas the AmORESTES sequences are expected to cover more central cDNA regions.
Even though the total number of ESTs available for Apis mellifera is still low when compared to established genomic model organisms, we performed an across genome analysis with the set of 3,408 honey bee contigs. This involved sequential BLASTX searches, using the honey bee sequences as query entries against protein databases of Drosophila melanogaster, Anopheles gambiae, Caenorhabditis elegans, human, protozoan and fungal origin. With this selection of organisms we intended to extract information on the percentage of genes that Apis shares (i) with all organisms, (ii) with animals, (iii) with different sets of metazoans, (iv) and exclusively with insects. The cutoff E-value in these comparisons was set at 10-6, as used in comparisons of similar nature , and the representation of the respective putative orthologs was listed across taxonomic levels.
Since deep-level phylogeny relationships within the bilateria are still a matter of debate, we separated our dataset according to the two prevalent hypotheses. The traditional view clusters arthropods within the coelomate clade. In our set of genomes, this tree architecture would be represented by genes shared between insects and the human genome. The alternative, more recently proposed hypothesis joins arthropods with nematodes to form an ecdysozoan clade . The result of our comparison, which places emphasis on shared genes and not on the frequency of gene losses, is more consistent with the traditional view, since the coelomate clade is represented in this analysis with almost five times more shared genes than the ecdysozoan clade.
To infer on functional aspects within this pattern of genes that different clades appear to have in common we performed a Gene Ontology classification on biological process. In the set of Apis ESTs that stands for genes putatively shared with all organisms, the majority was classified as having a role in metabolism, and thus can be considered to represent basic functions. In contrast, the majority of Apis ESTs that are shared within the insect clade was represented in the biological process categories of cell growth and/or maintenance and cell communication. The corresponding insect-specific genes are therefore supposedly involved in more specialized functions. A similar conclusion can be reached from the micro- and macroarray analyses of transcripts detected in adult honey bee workers performing different tasks during their adult life cycle [6, 7].
Annotation and Gene Ontology characteristics of 11 honey bee EST contigs sharing significant similarity with mammalian but not with other vertebrate or invertebrate sequences. In all cases, the best match was with human proteins. For these 11 out of 23 contigs we could retrieve functional information.
Apis EST contig
E-score value a
GO, Biological process b
Human LOCUS ID GenBank annotation
NM_019116: ubiquitin binding protein
ubiquitin-specific protease domain
NM_182830: MAM domain
contactin 5; neural adhesion molecule
NM_013041: RAB3A interacting protein (rabin3)-like1
guanin nucleotide exchange factor domain
NM_182565: hypothetical protein MGC29814
TBP-associated factor 4; TATA box binding protein
NM_172374: interleukin 4 induced 1
NM_024707: gem (nuclear organelle) associated protein
spliceosomal snRNP biogenesis
regulation of physiological process
NM_138457: forkhead box P4
transcription factor activity
NM_004654: ubiquitin-specific protease 9
ubiquitin thiolesterase activity
NM_014006: PI-3-kinase-related kinase SMG-1
involved in nonsense-mediated mRNA decay
NM_014035: sorting nexing 24
intracellular signaling cascade
Finally, we found that 1,779 (52,2%) of the assembled EST contigs did not match with any sequence of the analyzed organisms. Such a large proportion of Apis- specific contigs is likely to be an overestimate. As noted in a previous study , this might be partly due to technical problems, such as, sequencing of cDNA inserts consisting mainly of 3'-untranslated regions, the presence of unspliced intron sequences, cDNAs with a negative reading frame, or chimaeric cDNAs. However, the major portion of the Apis-specific contigs may have become classified as species-specific due to their relatively short ORFs. We performed an ESTScan analysis http://www.ch.embnet.org/software/ESTScan.html on the Apis-specific contigs which detected ORFs in 56% of the assembled ESTs. These ORFs are, however, relatively short, with a mean ORF length around 280 bp. Short ORF length represents a notorious problem to alignment algorithms resulting in low match scores, and consequently, a more frequent classification of short ORF ESTs as species-specific transcripts. For the honey bee, this has been shown for the brain cDNA library where 84% of the ESTs with ORFs shorter than 450 bp were classified as species-specific, against 24% in the EST set that had ORFs larger than 450 bp .
In order to gain a general perspective on the representation of species-specific ESTs we also directed our attention to estimates obtained in whole-genome cross-species analyses. For instance, a figure of 18.6% of species-specific genes was ascertained for Drosophila melanogaster in a genome comparison which included Anopheles gambiae as the other insect representative . Based on this information, and taking advantage of a set of Drosophila melanogaster ORESTES, generated in a parallel project, we calculated the frequency of Drosophila-specific ORESTES sequences to obtain a more realistic estimate on Apis-specific genes in relatively small sets of ESTs. In this analysis, a set of 5,000 CAP3 assembled Drosophila ORESTES (409 contigs) was submitted to sequential BLASTX searches against protein databases of Drosophila melanogaster, Anopheles gambiae, Caenorhabditis elegans, human, protozoan and fungal, as described for the Apis contigs. For comparison, this same analysis was also performed with Apis ESTs, using separately 5,000 AmORESTES (486 contigs) and 5,000 AmNCB (632 contigs).
The figure of 40% Drosophila-specific genes obtained for our Drosophila ORESTES set can be directly set in contrast with the estimate of 18,6% species-specific genes reported in the Drosophila genome based study , and this would predict an overestimate factor of 2.15 for species-specific genes in the EST sets. When this factor is applied to the honey bee ORESTES, the 47% estimate for species-specific AmORESTES can thus be corrected to a more realistic figure of 22%. This estimate is in agreement with the results of Whitfield et al.  who observed that 24% of the honey bee genes represented by ESTs with ORFs larger than 450 bp did not have matches to any known protein sequences. This Apis-specific gene estimate is also in range when considering that the two dipteran species are thought to have separated from a common ancestor approximately 250 million years ago, whereas the postulated sister-group relationship of Hymenoptera and Mecopteroidea  suggests a pre-permian divergence, with a predicted separate lineage evolution of over 280 million years for honey bees and dipterans .
The generation of a relative small set of Open Reading frame ESTs (ORESTES) that match and complement the already existent Apis EST database shows that this approach is sufficiently robust and favorably complements other strategies, such as ESTs prepared from normalized cDNA libraries. Its inherent properties of detecting transcripts of low abundance and aligning with central regions of transcripts [2, 3] also make it a suitable tool in searches for novel honey bee genes and their annotation in parallel with ongoing genome sequencing projects. Furthermore, the genome comparisons performed in this and other studies [1, 11] highlight that the elevated number of putative Apis-specific genes will still require extensive transcriptome sequencing for high quality genome annotation, and will play an important role in the question of insect genome organization and model systems in comparative studies .
Biological samples and RNA extraction
Samples of the four major stages of the honey bee life cycle were collected from Apis mellifera colonies (Africanized hybrids) kept in the experimental apiary of the Dept. Genetics, Univ. São Paulo, Campus Ribeirão Preto, Brazil. Each embryo sample contained approximately 300 eggs retrieved from a frame on which the queen had been caged for up to 72 hours. This assured that we covered the entire embryonic period. The larval sample was a representation of all five instars and included also spinning-stage larvae. Prepupae and pupae, including white-eyed, pink-eyed, brown-eyed and pigmenting pupae, were pooled into the pupal samples. For the adult sample we collected newly emerged bees, a random sample of hive bees (picked from a brood frame), and returning foragers. All these samples were snap frozen in liquid nitrogen. Total RNA was isolated using TRIzol reagent (Invitrogen). The lipid-rich larval and pupal samples required two additional extraction steps with phenol/chloroform and chloroform to obtain RNA of adequate purity.
In the case of Drosophila melanogaster, dechorionated embryos, larvae plus prepupae and pupae, as well as adult flies were collected from an isogenic y, w1118 stock of Drosophila melanogaster. These were immediately frozen in liquid nitrogen and stored at -80°C until use. Total RNA was extracted with TRIzol, as described for Apis mellifera.
Generation of Open Reading frame ESTs (ORESTES)
Specific primers used to assess quality and absence of DNA contaminants of the RNA samples, and randomly selected primers used to generate cDNA profiles.
actin F (Apis)
5' AGCTATGAACTTCCAGATGGT 3'
actin R (Apis)
5' CCACATCTGTTGGAAGGT 3'
16S mitochondrial F (Apis)
5' TTATTCACCTGTTTATCAAAACAT 3'
16S mitochondrial R (Apis)
5' 'TATAGATAGAAACCAAYCTG 3'
16S mitochondrial F (Drosophila)
5' CCGGTCTGAACTCAGATCACGT 3'
16S mitochondrial R (Drosophila)
5' CGCCTGTTTAACAAAAACAT 3'
5' TTGGGGATCGTATGTAGTATG 3'
5' CACTTCAGGATCCCTTGTAAGC 3'
5' CCAACATTGAATTCTCTTTGAC 3'
5' CAATAACAATGAATTCCAGAATCTCG 3'
5' GCTTACAAGGGATCCTGAAGTGTTTCC 3'
5' GCAGGTAAACTCTACTCGAGTTACG 3'
5' CCAGGATGTTTGGGTGATGTA 3'
5' TCATGCAACATCATCTGCTCC 3'
5' CTAACCCAAGACATGACATTC 3'
5' AAAGCTCGGAGGCAGCGAACT 3'
5' CATGTGGCTGCAAGGCAAAGC 3'
5' GCCCAATTGCAGTTGAGTGAT 3'
5' GTCTTGATGATGGTGAGAGTC 3'
5' TCATGTCCAACCCTAAGTCAC 3'
5' TCCAGCAGGCCTTGCAGAAAT 3'
5' TATCTCAAACGATCGAGAGAC 3'
5' GCACATCTATTACCAGCTTTG 3'
5' TTTCAAATGGCCACGGGAC 3'
5' GCACATTTATGACCATTCTCG 3'
5' AGAAACAACTCCAGGGGCCTG 3'
5' CTACCCAAAGCAGAAACCCCA 3'
5' CCAAAACCATCCTTGACAACA 3'
5' TGATGAGGTCCTTCACGGTG 3'
5' CGGAATTCACCAGATTTGAACAGAAGAG 3'
5' AACTGCAGTTAACCAGATTTGAACAGAAA 3'
5' CCGCTAGCATGTCCCCTATACTAGGTTA 3'
5' CGCTCCCGCTGTTTACTCT 3'
5' GACCGCTCCTCCAACTAACC 3'
5' CCGGCCCACCTCTTCTACTA 3'
5' TCTCTTTATGGCAAGACTTACG 3'
5' TCCTTAGCAACCATTAATCTGG 3'
5' GCCTATCTACTTCAGTGATTTCT 3'
5' ATCCAAGGTTCTCCCAATA 3'
ORESTES profiles were generated according to Dias-Neto et al. . Briefly, aliquots of 15 ng of purified mRNA were subjected to reverse transcription reactions utilizing SuperScript II Reverse Transcriptase (Invitrogen) and a set of randomly selected primers (Table 4). First-strand cDNA synthesis occurred at 37°C for 60 min in a total volume of 20 μl. The products of this reaction were diluted 1:5 in water and stored at -20°C. The cDNAs contained in 1 μl of each diluted RT-product were then amplified by PCR using the same or a single alternative random primer in a PCR mix (Ready-to-Go PCR bead, Amersham Biosciences). The amplification protocol consisted of an initial step at 75°C for 5 min, followed by a 45 cycles touchdown series (95°C for 30 s, a gradually decreasing annealing temperature from 66 to 44°C lasting 10 s per step and a decrease of 2°C per step, 72°C for 1 min), and a final extension reaction at 72°C for 7 min.
Aliquots of the PCR products (3–5 μl) were run on 1% agarose gels and stained with ethidium bromide. From profiles that presented near-even smears we excised two sets of amplification products, one covering a size range from 300 to 700 bp and a second one from 700 to 1500 bp. For cloning, these were extracted from the agarose gels (QIAquick Gel Extraction kit, Qiagen) and ligated into pUC18 (SureClone Ligation kit, Amersham Biosciences) for transformation of competent E. coli DH5α-cells by heat shock. Bacteria were grown in 2 × YT medium before aliquots were plated on 2 × YT agar containing ampicillin.
Blue-white selected positive colonies were picked, grown overnight in 2 × YT medium in 96-well plates, and used as templates for PCR using vector primers (M13 forward and reverse). An aliquot of each amplification product was analyzed on a 1% agarose gel before another 1 μl aliquot was submitted to DNA sequencing using standard protocols of the DYEnamic™ ET Terminator kit (Amersham Biosciences). The reaction products were analyzed in a MegaBACE™ 1000 automated sequencer. Only profiles with more than 80% positive PCR reactions were sequenced.
After passing through the Base Caller Cimaron 1.53 Slim Phredfy (insert size > 100, "N" nucleotides less than 20%, and "N" repetitions of less than 6 nucleotides) and ScoreCard procedure (MegaBACE™) to check sequences quality, reads that were larger than 100 nt were submitted to an automated protocol for data analysis (Gene Annotation Pipeline) of the Apis mellifera or Drosophila melanogaster ORESTES. The protocol consisted of the following steps: conversion of electropherograms (Phred, to formats .fasta, .phd and .qual), primer and vector detection and trimming (Cross_match) and masking of repeats (RepeatMasker). Validated fasta format sequences were then submitted to a general BLASTN search against GenBank entries for mitochondrial and rRNA, as well as bacterial and fungal RNA to detect and eliminate contaminant sequences.
For the Apis mellifera ORESTES, subsequent BLASTN searches were performed against the approximately 15,500 Apis mellifera EST sequences deposited in GenBank dbEST. In this case, significant E values were set at 10-30. Searches against the non-redundant protein database entries used the BLASTX option with E-values set at 10-10 as significance cut-off level.
CAP3 was used to clusterize the ORESTES sequences of both species. For Apis mellifera, the annotation of the 488 contigs was manually checked, giving preference to Drosophila sequences in the Unigene assignment. Subsequently, the contigs were batch submitted to a Gene Ontology procedure utilizing the FatiGO tools http://fatigo.bioinfo.cnio.es/. Clusterization of the Apis ORESTES contigs and singletons with the Apis mellifera ESTs deposited in GenBank dbEST was also performed using a CAP3 routine (standard parameters).
The production of Apis ORESTES was supported by grants from FAPESP (99/00719-6 and 98/142476) and that of Drosophila ORESTES received supported from a grant from Ludwig/FAPESP (99/03677-2). MCRC and NM were research fellows of FAPESP; KH received a CAPES/DAAD visiting professor fellowship, VV, DDA and JFS were supported by fellowships from FAPESP; FMFN, RMM, and MFRS received fellowships from CAPES and PMVP from CNPq. EDN is supported by ABADHS and CNPq grants. We thank Amélia G. Araujo, Fernanda Barbuzano, Cristiane A. Ferreira, Roberto Focosi Jr., Adriana A. Marques, Camila C.B.O. Menezes, Gislaine S.P. Pereira, Marlus C.O. Rocha, Anemari R.D. Santos, Cirlei A.V. Saraiva, Israel T. Silva and Benedita O. de Souza for their dedicated technical assistance.
- Whitfield CW, Band MR, Bonaldo MF, Kumar CG, Liu L, Pardinas J, Robertson HM, Soares B, Robinson GE: Annotated expressed sequence tags and cDNA microarrays for studies of brain and behaviour in the honey bee. Genome Res. 2002, 12: 555-566. 10.1101/gr.5302.PubMed CentralView ArticlePubMedGoogle Scholar
- Dias-Neto E, Correa RG, Verjovski-Almeida S, Briones MRS, Nagai MA, da Silva W, Zago MA, Bordin S, Costa FF, Goldman GH, Carvalho AF, Matsukuma A, Baia GS, Simpson DH, Brunstein A, de Oliveira PSL, Bucher P, Jongeneel CV, O'Hare MJ, Soares F, Brentani RR, Reis LFL, de Souza SJ, Simpson AJG: Shotgun sequencing of the human transcriptome with ORF expressed sequence tags. Proc Natl Acad Sci, USA. 2000, 97: 3491-3496. 10.1073/pnas.97.7.3491.PubMed CentralView ArticlePubMedGoogle Scholar
- Camargo AA, Samaia HPB, Dias-Neto E, Simao DF, Migotto IA, Briones MRS, Costa FF, Nagai MA, Verjovski-Almeida S, Zago MA, Andrade LEC, Carrer H, El-Dorry HFA, Espreafico EM, Habr-Gama A, Giannella-Neto D, Goldman GH, Gruber A, Hackel C, Kimura ET, Maciel RMB, Marie SKN, Martins EAL, Nobrega MP, Paco-Larson ML, Pardini M, Pereira GG, Pesquero JB, Rodrigues V, Rogatto SR, da Silva I, Sogayar MC, Sonati MDF, Tajara EH, Valentini SR, Alberto FL, Amaral MEJ, Aneas I, Arnaldi LAT, de Assis AM, Bengtson MH, Bergamo NA, Bombonato V, de Camargo MER, Canevari RA, Carraro DM, Cerutti JM, Correa MLC, Correa RFR, Costa MCR, Curcio C, Hokama POM, Ferreira AJS, Furuzawa GK, Gushiken T, Ho PL, Kimura E, Krieger JE, Leite LCC, Majumder P, Marins M, Marques ER, Melo ASA, Melo M, Mestriner CA, Miracca EC, Miranda DC, Nascimento A, Nobrega FG, Ojopi EPB, Pandolfi JRC, Pessoa LG, Prevedel AC, Rahal P, Rainho CA, Reis EMR, Ribeiro ML, da Ros N, de Sa RG, Sales MM, Sant'anna SC, dos Santos ML, da Silva AM, da Silva NP, Silva WA, da Silveira RA, Sousa JF, Stecconi D, Tsukumo F, Valente V, Soares F, Moreira ES, Nunes DN, Correa RG, Zalcberg H, Carvalho AF, Reis LFL, Brentani RR, Simpson AJG, de Souza SJ: The contribution of 700,000 ORF sequence tags to the definition of the human transcriptome. Proc Natl Acad Sci, USA. 2001, 98: 12103-12108. 10.1073/pnas.201182798.PubMed CentralView ArticlePubMedGoogle Scholar
- Verjovski-Almeida S, DeMarco R, Martins EA, Guimaraes PE, Ojopi EP, Paquola AC, Piazza JP, Nishiyama MY, Kitajima JP, Adamson RE, Ashton PD, Bonaldo MF, Coulson PS, Dillon GP, Farias LP, Gregorio SP, Ho PL, Leite RA, Malaquias LC, Marques RC, Miyasato PA, Nascimento AL, Ohlweiler FP, Reis EM, Ribeiro MA, Sa RG, Stukart GC, Soares MB, Gargioni C, Kawano T, Rodrigues V, Madeira AM, Wilson RA, Menck CF, Setubal JC, Leite LC, Dias-Neto E: Transcriptome analysis of the acoelomate human parasite Schistosoma mansoni. Nat Genet. 2003, 35: 148-157. 10.1038/ng1237.View ArticlePubMedGoogle Scholar
- Aguinaldo AMA, Turbeville JM, Linford LS, Rivera MC, Garey JR, Raff RA, Lake JA: Evidence for a clade of nematodes, arthropods and other moulting animals. Nature. 1997, 387: 489-493. 10.1038/387489a0.View ArticlePubMedGoogle Scholar
- Kucharski R, Maleszka R: Evaluation of differential gene expression during behavioral development in the honeybee using microarray and northern blots. Genome Biol. 2002, 3.2: research0007.1-9-[http://genomebiology.com/2002/3/2/research/0007]Google Scholar
- Tsuchimoto M, Aoki M, Tajkada M, Kanou Y, Sasagawa H, Kitagawa Y, Kadowaki T: The changes of gene expression in honeybee (Apis mellifera) brains associated with ages. Zool Sci. 2004, 21: 23-28. 10.2108/zsj.21.23.View ArticlePubMedGoogle Scholar
- Zdobnov EM, von Mering C, Letunic I, Torrents D, Suyama M, Copley RR, Christophides GK, Thomasova D, Holt RA, Subramanian GM, Mueller HM, Dimopoulos G, Law JH, Wells MA, Birney E, Charlab R, Halpern AL, Kokoza E, Kraft CL, Lai ZW, Lewis S, Louis C, Barillas-Mury C, Nusskern D, Rubin GM, Salzberg SL, Sutton GG, Topalis P, Wides R, Wincker P, Yandell M, Collins FH, Ribeiro J, Gelbart WM, Kafatos FC, Bork P: Comparative genome and proteome analysis of Anopheles gambiae and Drosophila melanogaster. Science. 2002, 298: 149-159. 10.1126/science.1077061.View ArticlePubMedGoogle Scholar
- Kristensen NP: The phylogeny of hexapod "orders". A critical review. Z Zool Syst Evol Forsch. 1975, 13: 1-44.View ArticleGoogle Scholar
- Crozier RH, Crozier YC: The mitochondrial genome of the honey bee Apis mellifera: Complete sequence and genome organization. Mol Biol Evol. 1993, 6: 399-411.Google Scholar
- Lobo NF, Ton LQ, Hill CA, Emore C, Romero-Severson J, Hunt GJ, Collins FH: Genomic analysis in the sting-2 quantitative trait locus for defensive behavior in the honey bee, Apis mellifera. Genome Res. 2003, 13: 2588-2593. 10.1101/gr.1634503.PubMed CentralView ArticlePubMedGoogle Scholar
- Evans JD, Gundersen-Rindal D: Beenomes to Bombyx: future directions in applied insect genomics. Genome Biol. 2003, 4: 107-110. 10.1186/gb-2003-4-3-107.PubMed CentralView ArticlePubMedGoogle Scholar
- Al-Shahrour F, Diaz-Uriarte R, Dopazo J: FatiGO: a web tool for finding significant associations of Gene Ontology terms with groups of genes. Bioinformatics. 2004, 20: 578-580. 10.1093/bioinformatics/btg455.View ArticlePubMedGoogle Scholar
This article is published under license to BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.