The complete EpapGV genome [GenBank: JN408834] was covered 34 times by 454 sequencing. It consists of 119,082 bp in good agreement with the previous estimate of 120.1 kbp based on restriction mapping 19 . Betabaculoviruses have AT-rich genomes ranging between 54.7% (CpGV) and 67.6% (CrleGV). The AT content of EpapGV genome is 58.5%. However, no correlation between these data and biological properties has been found thus far. Analysis of the EpapGV genome sequence led to the identification of 133 putative protein coding genes. The search was restricted to open reading frames starting with a methionine codon, coding for polypeptides of at least 50 amino acid residues (aa) and minimal overlapping of adjacent ORFs. This information comprises 90.94% of the nucleotide sequence (Additional File 1). The adenine of the granulin start codon was designated nucleotide 1 and the sequence was numbered in the direction of granulin gene transcription, which defined the clockwise orientation of the circular genome map 20 . The putative ORFs were numbered sequentially in this orientation. Seventy-two ORFs were in the same orientation as the granulin ORF, and sixty-one, in the opposite. EpapGV DNA sequence was searched for promoter motifs 150 bp upstream of the starting codon of each ORF. Early promoter motifs including TATA box (TATAWAW, TATAWTW, TATAW) in conjunction with CAKT initiator sequence (INR) 21 were found in the upstream regions of 26 ORFs; 64 ORFs had a late INR motif DTAAG 22 and 11 ORFs had both early and late elements.

The EpapGV genome contains the 31 core genes present in all baculoviruses. The genes were also classified according to their presence in different genera 23 24 (Figure 1).

A distinct feature of the EpapGV genome is that the core gene alkaline nuclease (alk-exo, epap119) is fused in frame with the helicase-2 ORF (epap120). This fusion gene codes for an 886 aa polypeptide with the first 383 residues homologous to Alk-Exo and the last 456 to Helicase-2. A 47 aa-long intervening polypeptide of unknown origin and without significant sequence similarity to any protein in GenBank was found between the Alk-Exo and the Helicase-2 regions. The intervening polypeptide may act as a low-structure linker between Alk-Exo and Helicase-2 such that both enzyme domains could fold as if they were independent polypeptides retaining their respective functions (Figure 2 and Additional files 2 and 3). Although this region was confirmed by resequencing, it will be important to study this genomic region in alternative isolates of EpapGV and determine the transcription and translation products in infected larvae. All GVs contain these two genes in the same order, but there are no reports of a fusion. Fusion genes seem to be extremely rare in baculovirus genomes, but there is one report of fused genes encoded by Spodoptera litura NPV: the ubiquitin ORF is fused in frame with gp37 and the fusion protein is proteolytically processed 25 .

For two EpapGV ORFs (epap10 and epap130) the BlastP search found homologues in only one member of the Baculoviridae. Epap10 is preceded by early and late promoter motifs and codes for a 90 aa protein that shares 34% amino acid identity with a 88 aa protein encoded by eppo28 of Epiphyas postvittana NPV. This ORF was reported to be unique to EppoNPV and has an early promoter motif 26 . Epap130 codes for a 77 aa protein that matched a 56 aa protein of Spodoptera litura GV (spli32) with 38% sequence identity according to ClustalW alignment.

ORFs 10, 46, 54, 55 and 105 were found to have homologues in alphabaculovirus genomes but not in other betabaculoviruses. Epap46 is a 306 aa long protein that seemed to be homologous to Spodopera exigua MNPV ORF 30 by BlastP search (E = 0.07) although they have very low amino acid identity (12%). Epap54 (148 aa) and epap55 (157 aa) are both similar to ORF 3 of Adoxophyes honmai NPV (AdhoNPV) and ORF 3 of Adoxophyes orana NPV (AdorNPV). Epap55 shares 32% identity with the N-terminal portion of AdorNPV ORF 3 and Epap54 is homologous to the C-terminal region of ORF 3 of AdorNPV (34%) and AdhoNPV (34%). GV homologues of epap54 and epap55 were found only in AgseGV (Additional File 1). Epap105 is similar to ac63 of Autographa californica MNPV (AcMNPV); their predicted proteins are 28% identical. Its homologue in Bombyx mori NPV (BmNPV), bm51, was reported to be a structural gene associated with the budded virus (BV) envelope 27 , but its deletion resulted in a virus with a phenotype similar to the wild type indicating that it might be a nonessential gene 28 .

Epap24 codes for a 388 aa long protein that is highly similar to ORF 21 of Cryptophlebia leucotreta GV (CrleGV) according to BlastP search (E = 4E-05). Crle21 is a 308 aa predicted protein reported to be similar to Se43 10 . These proteins share a protein motif of the DUF1383 superfamily. They have homologues in all alphabaculoviruses 28 and studies conducted with a deletion mutant of the homologue in AcMNPV (ac18) indicated that it is not essential for viral replication both in vitro and in vivo, but it may play a role in efficient virus infection in Trichoplusia ni larvae 29 . Homologues of epap24/crle21 were not found in the rest of the granuloviruses.

EpapGV DNA codes for the RNA polymerase subunits lef-4 (epap91), lef-9 (epap112), lef-8 (epap122) and p47 (epap63), lef-5 (epap82) and vlf-1 (epap101), present in all baculoviruses. Additional genes related to the transcription process found in all lepidopteran baculovirus were also detected in the EpapGV genome: 39 k/pp31 (epap56), lef-6 (epap74), lef-11 (epap57) (present in gammabaculoviruses) and pk-1 (epap6). Lef-10, involved in late transcription and present in most alpha- and betabaculoviruses, was also found in EpapGV genome (epap128). Of the baculoviral early transcription genes ie-0, ie-1, ie-2 and pe38, only ie-1 (epap35) is present in all GVs and pe38 was found in CpGV, CrleGV, PhopGV and PrGV.

Genes involved in DNA replication that belong to the core group were found in EpapGV genome: dnapol (epap106), lef-1 (epap68), lef-2 (epap41) and helicase-1 (epap85).

In addition, other genes that belong to this category and were found in EpapGV and in other lepidopteran baculoviruses are dbp (epap75) (also present in gammabaculoviruses), lef-3 (epap108), ie-1 (epap35), me53 (epap133) and ac38 (epap65). A lef-7 homologue was found in a BlastP search restricted to baculoviruses: the protein encoded by epap36 has a match with PsunGV LEF-7 (E = 0.54). This protein was demonstrated to be a baculoviral replication enhancer in AcMNPV 30 and BmNPV 31 . Homologues of this gene are present in group I NPVs, 3 group II NPVs and 3 GVs (XcGV, HearGV 28 and PsunGV).

EpapGV encodes a DNA ligase (epap115) as do other members of the Betabaculovirus genus and three NPVs of group II (LdNPV, LyxyNPV and OrleNPV). This gene seems to be linked to the presence of a second helicase, helicase-2 (epap120) 32 , which is fused with alk-exo in EpapGV, but not in the rest of the baculovirus genomes sequenced to date.

EpapGV genome contains all the structural genes corresponding to the core group as well as the lepidopteran baculovirus genes. The structural core group genes are: p6.9 (epap81), vp39 (epap92), vp1054 (epap129), vp91 (epap96), gp41 (epap99), odv-ec43 (epap43), odv-e18 (epap29), p74 (epap59), pif-1 (epap69), pif-2 (epap47), pif-3 (epap38); pif-4 (epap84); pif-5/odv-e56 (epap27) and the recently discovered pif-6 33 (epap109). Lepidopteran-specific baculovirus structural genes include granulin (epap1); 25 k-fp (epap113); odv-e25 (epap86); bv/odv-c42 (epap80), the last two are also present in gammabaculoviruses. F-protein (epap14) is the only gene shared by alpha-, beta- and deltabaculoviruses. EpapGV contains 42 of the 47 proteins found in the occlusion derived virus (ODV) of PiraGV 34 . Five of these ORFs were only found in betabaculoviruses: epap48, epap94, epap95, epap123 and epap126.

In addition to core gene alk-exo (Epap119), some other auxiliary genes were found in EpapGV genome. Viral ubiquitin (epap52) is present in GVs and all group I alphabaculoviruses. Cathepsin (epap31) and chitinase (epap32) were found in some GVs and in most alphabaculoviruses. These genes are responsible for the liquefaction of the host in the final stage of infection 35 36 . Their activity is readily apparent in E. aporema larvae infected with EpapGV. There is also a gp37 (epap30) homologue, which is present in some GVs and most NPVs. GP37 is homologous to the entomopoxvirus (EPV) fusolin which was shown to form spindle-like structures. These spindles enhance the peroral EPV infection by contributing to disruption of the peritrophic membrane 37 . EpapGV gp37 gene has been characterized and demonstrated to be glycosylated 38 . The EpapGV genome includes three fibroblast growth factor homologues: fgf-1, -2 and −3 (epap70, epap118 and epap131, respectively).

The three fgf genes are present in all sequenced GVs but fgf-2 is also present in all alphabaculoviruses. It is thought to be implicated in the virus dissemination in the insect host 39 40 . Epap58 encodes a superoxide dismutase homologue (sod) which is widely distributed in baculovirus. Its potential role is still unknown and controversial 41 .

EpapGV also possesses two iap genes (inhibitors of apoptosis), iap-3 (epap11) and iap-5 (epap111). Iap-5 is only present in betabaculovirus whereas iap-3 is also present in some NPVs. No p35 homologue was found.

The number of genes considered to be GV-specific has changed in the literature and will be more accurate when more complete genome sequences become available. These genes could be the basis to the differences between granuloviruses and nucleopolyhedroviruses. Taking into account the analyses presented by Lange et al. 10 , Wormleaton et al. 7 , Escasa et al. 9 , Van Oers & Vlak 42 , Miele et al. 23 and the present report, a set of 19 genes has been identified in betabaculovirus genomes which were never found in alpha-, gamma- or deltabaculoviruses (Figure 1, Additional file 1). These are EpapGV ORFs 7, 8, 17, 21, 22, 25, 37, 40, 43, 44 (metalloproteinase), 62, 70 (fgf-1), 73, 94, 95, 110, 111 (iap-5), 126 and 131 (fgf-3).

Other genes formerly considered as part of the GV-specific set, were dismissed from the list in the present report: CpGV ORFs 30, 32, 45, 50, 56, 77, 82, 119, 121, 122 and 136. All, except cp27, 56, 77, 121 and 136, have homologues in EpapGV but they are absent in some other GV (see Additional file 1).

Seventeen ORFs appear to be unique to EpapGV compared to the rest of the members of Baculoviridae (ORFs 4, 9, 12, 16, 18, 19, 20, 23, 49, 51, 60, 64, 72, 89, 104, 114 and 116). Epap4 codes for a 144 aa long protein with a conserved motif (COG5152) in its N-terminal region. This motif is an uncharacterized conserved domain that contains RING and CCCH-type Zn-fingers 43 . An early promoter motif was found 150 nt upstream epap4 ORF. Epap9 encodes an 81 aa long polypeptide and has no significant BlastP matches. The upstream region contains a GATA motif (TGATAG) and two TATAWAW early promoter elements, but no CAKT INR. Epap12 codes for a 90 aa protein which shares 23% identity and 43% similarity with a small portion of a 2123 aa protein of Drosophila ananassae (XP_001953497); however, no speculation on function can be made. Epap16 gives no significant BlastP hit, and has early promoter elements upstream of the first ATG (TATAW + 3 CATT elements). Something similar happens with Epap18 which codes for a hypothetical 76 aa protein and a TATAW element upstream. Epap19 (94 aa) has no significant BlastP hits and has elements of an early promoter. Epap20 (422 aa) has no significant BlastP hits and shows elements of a late promoter. Epap23 codes for a hypothetical protein of 197 aa with no significant similarity with any protein of the GenBank under the control of a putative late promoter and a GATA motif (TGATAG).

Epap49 codes for the longest hypothetical protein of EpapGV genome (1465 aa). As it lacks characteristic promoter elements and exhibits no similarity with other baculovirus genes it is difficult to predict if it is actually transcribed. Epap49 is located between the conserved genes pif-2 (epap47, core gene) and epap50 (homologue to cp52). It is worth mentioning that at least in two GVs (HearGV and ChocGV) a similar situation emerged in the same locus. Although a 1144 aa ORF with 27 leucine zippers was initially found in ChocGV, it was not considered a coding sequence but a non-hr ori-like region instead; the speculation was based upon its very high AT content (81%), the lack of homology with baculovirus ORFs, and the possibility of sequencing errors (for further details see Escasa et al., 9 ). In contrast, the 1279 aa ORF in HearGV (hear44) was considered a coding sequence resulting from a fusion of the homologues xc47 and xc48 of Xestia c-nigrum GV 12 .

Epap51 codes for a 69 aa peptide under the control of an early promoter and showed no significant matches in BlastP search. Two GATA motifs (TGATAT and AGATAG) were also found in the region upstream its ATG. Epap60 (582 aa) shows no significant hits with any protein in the GenBank. TATATAA and TATAA motifs were found upstream the ATG, but without the initiator sequence CAKT, characteristic of early promoters. Epap64 codes for a thymidylate kinase (described below). Epap72 codes for a 61 aa peptide with no significant matches in the GenBank under the control of a putative late promoter which is overlapped with a GATA motif (AGATAAG). Epap89 predicted protein (86 aa) did not have significant BlastP hits either and lacks known promoter motifs except for a TATAAAA sequence 86 nt upstream its ATG overlapped with a GATA motif. Similarly, ORFs 104 (63 aa), 114 (162 aa) and 116 (51 aa) contain TATA box-like motifs upstream the initial ATG and show no significant BlastP hits. Epap114 also presents a ATAAG sequence, and Epap116 a GATA motif (AGATAA).

Genes coding for enzymes involved in nucleotide metabolism have been reported in baculovirus genomes. Ribonucleotide reductase (RNR) catalyses the reduction of ribose in ribonucleotide diphosphates to yield deoxyribonucleotides, the building blocks of DNA 44 . In most eukaryotes the active RNR is a tetrameric complex made up of homodimers of two subunits coded by genes rr1 and rr2, which have been also found in some NPVs and 4 GVs: CpGV, AgseGV, PhopGV and EpapGV (epap2 and epap3). On the other hand, dUTPase catalyses the dephosphorylation of dUTP to yield dUMP. As dUTP can be mutagenic if incorporated in DNA, this enzyme helps to keep levels of dUTP low and prevents its incorporation in DNA, in lieu of dTTP. This gene is present in some NPVs and in the betabaculoviruses AgseGV, SpltGV and EpapGV (Epap13). The presence of rr1, rr2 and dutpase appears to be linked in the genomes of OpMNPV, SpltNPV, SeMNPV and LdMNPV 32 . This linkage appears in the betabaculoviruses AgseGV and EpapGV but not in SpltGV. Both enzymes participate in the pathway of de novo dTTP biosynthesis (Figure 3a).

EpapGV codes for a novel enzyme in the family Baculoviridae, which also takes part in this pathway: epap64 codes for a predicted 224 aa protein homologous to thymidylate kinase, also known as thymidine monophosphate kinase (TMPK), that catalyses the phosphorylation of dTMP to produce dTDP. BlastP hits included different eukaryotic organisms and several viruses representing the families Poxviridae (Variola Virus), Iridoviridae (Invertebrate iridescent virus 6, II6), Herpesviridae (Cyprinid Herpes 3, CyHV3), Nimaviridae (White Spot Syndrome Virus, WSSV) and Asfaviridae (African swine fever virus, ASFV) that were used in the ClustalW alignment with EpapGV TMPK (Figure 3b). EpapGV TMPK showed the highest identity (40%) with TMPK from the insect Drosophila ananassae and the least (22%) with TMPK from ASFV. The degree of identity with the other viruses was 35% (II6); 32% (Variola and Vaccinia); and 33% (WSSV). Besides Baculoviridae, other viral families that encode nucleotide metabolism genes include Herpesviridae, Poxviridae and Asfaviridae. The alphaherpesvirus pyrimidine deoxynucleoside kinase, popularly known as thymidine kinase (TK) phosphorylates a wide range of nucleoside substrates, as well as TMP (TK + TMPK activity), and is responsible for the rise in the TTP pool characteristic of HSV-infected cells 45 . In poxviruses these TK and TMPK activities are present in separate enzymes as happens in cellular organisms. Vaccinia virus TMPK was found to be nonessential for virus replication in cultured cells and able to complement a tmpk- Saccharomyces cerevisiae mutant 46 .

The White Spot Syndrome Virus (WSSV; Nimaviridae) genome contains a mosaic gene that encodes a tk-tmpk fusion of both homologues, i.e. cellular-type thymidine kinase TK1 and cellular-type TMPK 47 . However, only TK activity, but not TMPK, could be demonstrated for WSSV TK-TMK 48 . TMPK substrate specificity was studied in vaccinia virus and it was found to phosphorylate dTMP, dUMP and, unlike human TMPK, dGMP as well 49 . EpapGV TMPK expression and substrate specificity, as well as its role in infection, remain to be elucidated.

A common feature in baculovirus genomes is the presence of nucleotide sequence repeats known as homologous regions (hrs). These regions function as enhancers of early gene transcription and are thought to play a role as origins of replication. They are characterized by tandem copies of sequence motifs that include an imperfect palindromic core. Although they present significant sequence similarity within a genome they are highly variable when compared between any two different species (Reviewed in 50 ).

In a first screening of the EpapGV genome for repeated sequences with Blast2seq we found two palindromic regions, of 128 bp (58352–58479) and 122 bp (116114–116235), respectively. Both sequences are very likely to exist in equilibrium between double stranded DNA and opposite hairpin-loops constituted by each complementary strand (hr10, -75.30 kcal/mol and hr16a, -86.50 kcal/mol, respectively) forming a cruciform-like structure (Figure 4a,b). Whether this feature is biologically relevant remains to be elucidated. Using these two sequences as profiles we searched the rest of the genome for similar sequences. Twenty-four AT-rich sequences of similar size were detected dispersed throughout the genome. The alignment of these sequences revealed the presence of conserved palindromes of about 58 bp that correspond with the central part of the two initially identified largest palindromes, with a mean of 70% AT content. The alignment of these shorter palindromes (Figure 4d) shows that they have an AT rich core flanked by 15 bp conserved inverted repeats (Figure 4d,e). This structure is similar to that of hrs found in all GVs (those sequenced to date) that infect other insects of the Tortricidae family (CpGV, CrleGV, AdorGV, ChocGV): palindromic sequences of about 63–76 bp characterized by conserved blocks of 13 bp located at both ends. These ends were found to be similar in sequence not only among the palindromes of a single genome but among the different genomes, as well 51 . EpapGV palindromic ends are similar to the consensus 13 bp sequence for these GVs, including PhopGV (Figure 4f). PhopGV, that infects a member of the Gelechiidae family but appears in the same clade as these GVs, was also found to have this kind of palindromic repeats, although their length was different (142–320 bp) 51 . Palindromes sharing these features were analyzed in infection-dependent replication assays in cells susceptible to CpGV infection. All 14 CpGV palindromes were found to act as origins of replication of plasmids in infection-dependent assays in Cydia pommonela cells 51 52 , whereas the hrs from other GVs tested in the same way did not show this ability, with the only exception of 2 palindromes of the most closely related CpGV virus, CrleGV 51 .

These palindromic sequences are found with a much greater frequency in EpapGV DNA compared to the other tortricid-specific GVs. EpapGV contains 26 palindromes (within 16 hrs), whereas the others have up to 17 palindromes as is the case of CrleGV 51 . In contrast, only four hrs (each one containing only 2–3 direct repeats) were reported in the most recently published betabaculovirus genome sequence (ClanGV; 18 ).

There seem to be some conserved locations for the hrs in GV genomes. For example, the region between sod and p74 and downstream of the CpGV ORF 5 51 .

It has been reported for AcMNPV that VLF-1 (a protein present in all the baculoviruses sequenced to date) binds with high affinity to cruciform DNA structures and it was suggested that this may play an important role in the replication/packaging process 53 . These cruciform structures, formed by the two largest palindromes or by the smaller ones interspersed in the EpapGV genome, may as well interact with VLF-1 and play a role in the replication or packaging.

In addition to the 26 palindromes mentioned above, there is a large structure consisting of 327 bp flanked by the 15 bp conserved ends predicted to form the secondary structures shown in Figure 4(a). This structure is located in the hr4 region (including hr4a and hr4b), an AT-rich sequence between ORFs 17 and 19. The sequence organization is depicted in Figure 4(c) showing ORF 18 within hr4b, which also contains two 31 bp direct repeats (A1, A2), and an intergenic region with a second pair of imperfect direct repeats of 79 and 72 bp (B1, B2), respectively. Interestingly, this region is located in the same relative position where a putative non-hr ori was described in CpGV spanning ORFs 24, 25 and 26 (which are absent in EpapGV) 11 and in CrleGV 54 .

Strong colinearity is observed in granulovirus genomes sequenced to date 7 10 11 51 . Baculovirus gene colinearity has been analysed mainly by Gene Parity Plot 55 . In this study we used the Artemis Comparison Tool to analyse the gene synteny of EpapGV compared to all other sequenced GVs and the type species of the Alphabaculovirus genus, AcMNPV. This tool enables to construct synteny maps through a tBlastX comparison between genomes, where inversions are easily detected as well as the different percentages of identities that correlate with different colour intensity. Figure 5 shows the conserved gene colinearity of all 13 sequenced GV genomes and the poorly conserved synteny between GVs and AcMNPV. Notably EpapGV differs from the rest of the GVs by a ca. 20 kb gene block inversion, as we noted previously in a physical map 19 .

Phylogenetic analysis based on 31 concatenated core genes of 58 baculovirus genomes was performed (Figure 6). The obtained cladogram reproduced the grouping of four genera recognized in the current classification of the family 1 . Division in two main groups of the Alphabaculovirus genus agrees with the group I and II. Two clades (Ia and Ib) previously described in group I by using concatenated amino acid sequences of the partial polh/gran, lef-8 and lef-9 genes 56 were also confirmed in our analysis.

As expected, EpapGV grouped in the Betabaculovirus genus. In previous reports it was observed that totricidae and noctuidae specific GVs tend to be in separated groups 56 57 . The cladogram obtained in this work confirms previous observations, and the additional complete genomes considered here allowed the division of betabaculoviruses in two well separated monophyletic clades as reported previously 23 . Clade “a” includes six species: PxGV, AgseGV, SpliGV, PsunGV and XcGV, which were isolated mainly from Noctuidae hosts. PxGV is the exception; its host belongs to the Plutellidae family. Clade “b” includes seven species: EpapGV, AdorGV, PhopGV, CpGV, CrleGV, PiraGV and ChocGV; five of them were isolated from Tortricidae, whereas PiraGV was isolated from Pieridae and PhopGV, from Gelechiidae. EpapGV seems to be the GV isolate closest to the common ancestor of Clade “b”. Both clades includes slow killing (type 1 GVs) and fast killing (type 2 GVs), reinforcing the concept of that this biological feature is not phylogenetically informative 57 .

EpapGV was originally isolated from a larva of the bean shoot borer Epinotia aporema collected in Oliveros (Santa Fe, Argentina) 5 . It was amplified allowing fourth instars to feed on artificial diet superficially contaminated with EpapGV occlusion bodies (OBs). Moribund larvae were collected and processed according to Parola et al. 19 : viral DNA was isolated from sucrose gradient purified OBs. Its integrity and identity was checked by restriction digestion and agarose gel electrophoresis.

EpapGV genomic DNA was sequenced with the 454 Genome Sequencer (GS) FLX™ Standard (Roche) at the Interdisciplinary Center for Biotechnology Research (ICBR), University of Florida (Gainesville, US). De novo assembly was generated on newBler assembler (GS FLX Data Analysis Software).

Open reading frames (ORFs) were identified using VectorNTI software (Invitrogen) and ORF Finder http://www.ncbi.nlm.nih.gov/gorf/gorf.html 58 . ATG initiated ORFs of at least 150 nt (50 aa) with minimal overlap were selected for further analysis. Homology searches were done using Blast 59 . Percentage identities between homologous genes were obtained by global alignments with ClustalW 60 using default parameters. Early (E) and late (L) Promoter motifs within 150 bp upstream of the putative ORFs were screened. E indicates the presence of a TATA-box (TATAW, TATAWAW, TATAWTW) with a CAKT mRNA start site 20–40 nucleotides downstream; whereas L denotes a DTAAG sequence 7 10 61 . Also GATA motifs WGATAR 62 and WGATAY 63 were searched for the unique genes.

Prediction of secondary structure of Alk-Exo_Helicase-2 fused protein was performed with the Jpred3 server 64 ; http://www.compbio.dundee.ac.uk/www-jpred/) using default parameters and single sequence submit. Actually, the prediction accuracy of Jnet (main Jpred3 algorithm) raised 81.5% in blind tests with soluble proteins. C-terminal end of Alk-Exo and N-terminus of Helicase-2 were selected on the basis of multiple alignments of the respective GV proteins.

Repeated sequences were searched first aligning EpapGV genome to itself through Blast2seq program from NCBI 65 . The first hit which corresponds to the 100% match of the complete genome was ignored and the following hits were used for further analysis. The consensus alignment obtained from two palindromes was used to find similiar sequences along the genome with the VectorNTI program (Invitrogen). The secondary DNA structure prediction of these sequences were performed in the Mfold server of The Vienna RNA website 66 . The alignment of all the palindromes found was performed with ClustalW algorithm with default parameters. The sequence logo of this alignment was carried out at the WebLogo server (http://weblogo.berkeley.edu/) 67 .

EpapGV genome was compared with other baculovirus genomes by constructing syntenic maps with the Artemis Comparison Tool (ACT) 68 (The Sanger Institute; http://www.sanger.ac.uk/resources/software/act/), using tBlastX program.

Phylogenetic analysis was performed using 31 core genes from 58 baculovirus genomes (Additional File 4) which were independently aligned using ClustalX program 69 , with the following parameters: Pairwise alignment (Gap Open Penalty = 10, Gap Extension Penalty = 0.1, protein weight matrix: Blosum 30); Multiple alignment (Gap Open Penalty = 10, Gap Extension Penalty = 0.05, protein weight matrix: Blosum series). Then a concatemer was generated by addition of the complete individual alignments and phylogeny was inferred using MEGA 5 program 70 with the following parameters: UPGMA; Bootstrap with 1000 replicates; gap/Missing data = complete deletion; Model = Amino (Dayhoff Matrix); patterns among sites = Same (Homogeneous); rates among sites = Different (Gamma Distributed); gamma parameter = 2.25. The obtained data was deposited in TreeBASE (http://purl.org/phylo/treebase/phylows /study/TB2:S12862).