![[Feedback]](/icons/banner/feedbackACT.gif){kind=icon}
![[For Subscribers]](/icons/banner/subscriberACT.gif){kind=icon}
![[Archive]](/icons/banner/archiveACT.gif){kind=icon}
![[Search]](/icons/banner/searchACT.gif){kind=icon}
![[Contents]](/icons/banner/tocACT.gif){kind=icon}
|
|
|
|
|
||
|
LETTER Gene Discovery in the Apicomplexa as Revealed by EST Sequencing and Assembly of a Comparative Gene Database1Department of Biology, 2Center for Bioinformatics, University of Pennsylvania, Philadelphia, Pennsylvania 19104, USA; 3Department of Genetics, University of Georgia, Athens, Georgia 30602, USA; 4Genome Sequencing Center, Department of Genetics, 5Department of Molecular Microbiology, Washington University School of Medicine, St. Louis, Missouri 63108, USA; 6Department of Veterinary Sciences, University of Kentucky, Lexington, Kentucky 40546, USA;7 Human and Animal Infectious Diseases, Merck Research Laboratories, Rahway, New Jersey 07065, USA; 8Veterinary Molecular Biology, Montana State University, Bozeman, Montana 59717, USA
Large-scale EST sequencing projects for several important parasites within the phylum Apicomplexa were undertaken for the purpose of gene discovery. Included were several parasites of medical importance (Plasmodium falciparum, Toxoplasma gondii) and others of veterinary importance (Eimeria tenella, Sarcocystis neurona, and Neospora caninum). A total of 55,192 ESTs, deposited into dbEST/GenBank, were included in the analyses. The resulting sequences have been clustered into nonredundant gene assemblies and deposited into a relational database that supports a variety of sequence and text searches. This database has been used to compare the gene assemblies using BLAST similarity comparisons to the public protein databases to identify putative genes. Of these new entries,  15%–20% represent putative homologs with a conservative
cutoff of p < 10–9, thus identifying many
conserved genes that are likely to share common functions with other
well-studied organisms. Gene assemblies were also used to identify
strain polymorphisms, examine stage-specific expression, and identify
gene families. An interesting class of genes that are confined to
members of this phylum and not shared by plants, animals, or fungi, was
identified. These genes likely mediate the novel biological features of
members of the Apicomplexa and hence offer great potential for
biological investigation and as possible therapeutic
targets. [The sequence data from this study have been submitted to dbEST division of GenBank under accession nos.: Toxoplasma gondii: BG657138–BG661027, BI921045–BI921090, BI946571–BI946588, BM003839–BM004582, BM039066– BM040645, BM131233–BM133172, BM174962–BM176879, BM188953–BM189923, BM271559–BM271694. Plasmodium falciparum: BI670521–BI670830, BI813842–BI816393, BI936022–BI936312, BM273300–BM276553. Sarcocystis neurona: BE574328, BE574347, BE574384, BE574386, BE574409, BE574465, BE574508, BE574543, BE574561, BE574633, BE574689, BE574694, BE574723, BE635418–BE636244, BE574288–BE574724, BF323572–BF324064, BM252128–BM253024, BM303125–BM305293. Eimeria tenella: AI755306–AI758088, AI759179–AI759181, AI759254–AI759304, AI759182–AI759253, AI759305–AI759387, AI759463–AI759546, AI759388–AI759462, AI759547–AI759621, BE027133– BE028807, BF023640–BF023711, BF023609–BF023639, BG235514–BG235880, BG413067–BG413336, BG466192–BG467045, BG515959–BG517044, BG560819–BG562379, BG724474–BG725148, BI895002–BI896127, BM305294–BM306971, BM321464–BM322026. Neospora caninum: BF248514–BF249435, BF716421–BF717094, BF823742, BF823805– BF823813, BF823743–BF824633, BG235070–BG235513.]
The generation of expressed sequence tags (ESTs) provides a rapid means of gene discovery from single-pass sequencing of randomly selected cDNAs. This approach has been particularly useful for complex, model genomes including human (Hillier et al. 1996  ), rat
(Scheetz et al. 2001), mouse (Marra et al. 1999b), fish
(Clark et al. 2001), and rice (Ewing et al. 1999). One of the primary
advantages of ESTs is that the identification of putative genes by
BLAST comparisons (Altschul et al. 1990) enables researchers to begin
biological analyses prior to the completion, or even initiation, of a
full genome sequence. EST sequencing is likely to make its greatest
impact on understudied genomes where little prior sequence data exists
and where full genome sequencing projects may not be undertaken in the
near future. Parasites provide such a group, and previous EST projects
have revealed the tremendous utility of this approach for gene
discovery (Reddy et al. 1993; Chakrabarti et al. 1994; Wan et al. 1995;
Ajioka et al. 1998; Manger et al. 1998b; Howe 2001). EST
sequencing allows not only the rapid identification of abundantly
expressed genes, it also provides data sets for informing phylogenetic
analyses, examining strain diversity, and exploring developmentally
regulated genes. Tools for recognizing and combining ESTs generated
from the same gene into nonredundant assemblies in silico have recently
been refined, improving the chances for establishing gene identities.
Such identities, although only putative, enable rapid analysis of gene
function, thus greatly facilitating traditional research approaches.
To further the process of gene discovery in protozoan parasites, we
have undertaken large-scale EST sequencing projects for several
apicomplexan parasites. The Apicomplexa is an ancient phylum of
One of the major drawbacks of EST sequencing is the large number of database entries that are submitted separately to dbEST without extensive annotation. This complicates the problem of establishing which ESTs belong to a given gene, and whether similar ESTs belong to the same or closely related genes. Therefore, in addition to generating new ESTs from these organisms, we have clustered and assembled the resulting sequences into RNA consensus sequences and created a gene database that provides a variety of features for comparative analyses. Herein, we describe the creation of this database, illustrate several of its important features, and highlight several major features of gene content and expression in the Apicomplexa.
General Strategy for Rapid Generation of ESTs cDNA libraries were initially screened to ensure that they contained a high percentage of inserts (  95% of 500 bp in size) and a
reasonable diversity of sequences based on analyses of 50–100 clones
per library. Sequences were subsequently obtained by single 5' reads
that were generated from randomly chosen cDNA clones. The numbers of
ESTs generated for the organisms studied here are shown in Table
1. Sequences were analyzed for high
quality, trimmed to remove vector sequences, and submitted to the dbEST
division of GenBank, providing they met the criteria outlined in
Methods. In general, the success rate, defined as the percentage of
sequences that passed the submission criteria, ranged from 70%–80%,
depending on the library in question. This strategy is rapid,
economical, and amenable to high-throughput production. Sequencing of
only the 5' ends of cDNAs has been used previously in the T.
gondii EST project (Wan et al. 1995; Ajioka et al. 1998; Manger et
al. 1998b), and proved valuable for detecting gene
similarities. In part, this is because of the early truncation of many
cDNAs, combined with the relatively short 5'- UTRs, such that the
regions sequenced tend to lie within the open reading frame, thus
facilitating gene identification. A similar strategy was used for all
of the projects described here. All of the EST sequences were initially
compared with the mRNA database, and similarities were annotated at the
time of submission.
Gene Assemblies and Generation of a Comparative Database There are inherent difficulties in using raw EST sequences to identify genes because they are incomplete, highly redundant, and error prone. To alleviate these problems, we clustered the ESTs separately for each of the organisms shown in Table 1 to form consensus sequences. These clustered sequences were further grouped with their corresponding mRNAs in GenBank to generate "assemblies," which represent consensus sequences for a given transcript within each organism. To better facilitate comparative analyses, we used a relational database called GUS (Genomics Unified Schema; Davidson et al. 2001 ). This apicomplexan EST database comprises the separate
assemblies from each species along with their annotations and
information on the EST sources. The vast majority of these sequences
were newly generated in the present project, although significant
numbers of T. gondii ESTs have been reported on previously
(Wan et al. 1995; Ajioka et al. 1998; Manger et al. 1998b),
or in the case of P. falciparum were deposited into dbEST by
others (Reddy et al. 1993; Chakrabarti et al. 1994). To make these data
fully accessible and allow other researchers to ask their own
questions, we have made the data and results available through a
user-friendly Web interface called ApiESTDB
(http://www.cbil.upenn.edu/paradbs-servlet/). The Web site was
generated by Java Servlets and Perl programs previously developed for
the AllGenes database (http://www.allgenes.org). ApiESTDB can be
queried by information on the constituent ESTs (e.g., library,
accession), key word searches on similarities to known genes and
protein domains, and BLAST searches of a user-defined sequence. The
database also supports Boolean queries that can be built by users on
top of the available queries and is able to retain the history of a
series of searches performed, whose results can be selectively combined
by users. For each selected assembly, a summary page shows graphical
displays of similar protein sequences and protein domains from BLAST
results as well as links to alignment and library breakdown of the
constituent ESTs (Fig. 2).
Identification of Genes Once compiled, the assemblies for each organism were compared separately to SWISS-PROT/PIR using BLASTX similarity searches to identify previously known genes using a cutoff of 98% identity (referred to as "known genes" in Table 1). These genes represent previously characterized, full-length cDNAs from each of the organisms. Notably, the numbers of previously known genes are extremely low for E. tenella (8), N. caninum (2), and S. neurona (0), reflecting the limited extent to which these organisms have been studied. Although more genes have been previously described for T. gondii (72) and P. falciparum (243), these still represent a small fraction of the expected number of total genes.
To find putative homologs corresponding to conserved genes that have
not been characterized in these parasites previously, the assemblies
were compared with the SWISS-PROT/PIR (excluding hypothetical or
unknown proteins) using BLASTX with a cutoff of
p-value < 10–9. A large number of statistically
significant similarities were detected for each of the species,
accounting for The search parameters used here for gene comparisons were quite selective, and it is likely that more sensitive searches would identify additional similarities. ApiESTDB allows investigators to perform a variety of self-directed BLAST searches and the standard parameters (E, S, matrix) can be set by the user. Consequently, the analysis performed here should be considered as a relatively selective assessment of gene similarities rather than an exhaustive analysis.
Redundancy
The redundancy inherent in cDNA libraries is beneficial in the generation of assemblies as overlapping ESTs from a single gene can be aligned and compiled to generate a consensus sequence in silico. Comparison of multiple ESTs from a given assembly is useful for identifying the boundaries of open reading frames, alternative splicing, and strain-specific polymorphisms. Putative alternatively spliced transcripts were identified by clustering the consensus sequences of assemblies with a minimum cutoff of 95% identity over >150 bp length and cross-linking related assemblies to each other. In the case of T. gondii, libraries were chosen to represent each of the three predominant genetic lineages (Howe and Sibley 1995 ), and
single-nucleotide polymorphisms are highlighted in the alignments for a
given assembly (see Fig. 2). Additionally, the source library for each
EST is retained, allowing easy determination of the relative abundance
in a given life cycle stage or strain type.
Abundantly Expressed Genes
Stage-Specific Gene Expression Apicomplexan life cycles involve a number of developmental adaptations that are designed to accommodate changes in the environment. These include changes between different hosts, either vertebrate or invertebrate, and between the host and extracellular environments (i.e., oocyst shedding). We analyzed the number of genes that are expressed in a stage-specific manner based on an arbitrary cutoff as well as statistical significance of their differential expression levels as reflected by abundance of ESTs at each stage (Table 2). A listing of these stage-specific assemblies can be found at: http://www.cbil.upenn.edu/paradbs-servlet/pub/SupTable2Stage.htm.In the case of P. falciparum, we obtained ESTs only from the asexual forms that replicate in blood cells and from gametocytes and did not include mosquito or liver stages. More than 80% of assemblies were stage-specific as judged by being fivefold different; however, only 11% of these were statistically significant (p  0.05). In the case of E. tenella, the two
predominate developmental stages are intracellular merozoites, that
replicate within the intestine, and oocysts, which survive in the
environment and contain sporozoites that are responsible for
transmission. In comparing these two different stages, again >80% of
assemblies were stage-specific by the fivefold criteria, whereas
20% of these were statistically significant (p 0.05).
In the case of T. gondii, we were able to generate cDNA
libraries from two of the major replicating forms, tachyzoites,
bradyzoites, and from oocysts, the extracellular form responsible for
environmental transmission. Approximately 2.3%, 5.6%, and 8.5%
assemblies were specific to each of these stages, respectively,
indicating that over all 16% of genes are expressed in a
stage-specific manner based on the p 0.05 cutoff (Table
2). Little is known about the molecular mechanisms of stage-specific
expression. However, these life cycle stages must survive in very
different environments, and presumably these changes in gene expression
underlie important biological adaptations that are specifically needed
in each of these niches. One caveat to the analyses presented here is
that the relative abundance of a given transcript in a particular
library may have been influenced by procedures in library construction.
More sensitive methods for examining stage-specific expression, such as
microarrays (Lashkari et al. 1997; Eisen and Brown 1999; Cummings and
Relman 2000) or SAGE (Velculescu et al. 1995), will be needed to
validate and extend these findings to other gene families. The
availability of an EST gene database will facilitate such analysis by
providing target genes for microarray studies, and by providing
sequences for comparison to SAGE tags.
Phylogenetic Comparison of Gene Homologies In classifying the orthologs identified by EST sequencing, it was of interest to determine the phylogenetic associations of the closest neighbor. Such information could be useful for establishing gene origins and for identifying genes with restricted phylogenetic distributions versus those that are widely dispersed. Thus, we further classified those assemblies that have homologs obtained from Table1 based on the phylogenetic origin of their putative orthologs (see http://www.ncbi.nlm.nih.gov/Taxonomy/taxonomyhome.html/; Table 3). Surprisingly, only 1%–2% of all assemblies have best non-self (not from the same organism) similarity to another apicomplexan. In part, this likely reflects the relatively small coverage of most of these genomes and underrepresentation of proteins from these organisms in the database. Approximately 3%–5% showed the highest similarity outside the phylum Apicomplexa to a plant gene, whereas 5%–10% were most similar to animals (Table 3). Previous studies have also highlighted the fact that many metabolic enzymes in T. gondii are plant-like (Dzierszinski et al. 2001 ). These plant-like genes may have originated from an algal
endosymbiont that is still represented by the remnant apicoplast
(Kohler et al. 1997; Roos et al. 1999). Similarities to archaea and
fungi were less common, and these may represent important examples of
convergence or of lateral gene transfer. One of the major implications
of the presence of plant-like genes, and the less common, but
nonetheless important, prokaryotic-like genes, is that they identify
metabolic pathways in apicomplexans that may be significantly different
from those of their mammalian hosts. Thus, these pathways represent
prime candidates for development of new therapeutics that might be
expected to offer a high degree of selectivity. It is possible to
further analyze such candidates in terms of putative function and
domain structure using ApiESTDB to design specific searches and
queries.
Identification of Genes Restricted to the Apicomplexa One of the primary reasons to undertake the present project was to identify genes that are restricted to the Apicomplexa, as these may mediate their unique biology. We performed two comparisons to identify apicomplexan-specific genes, dividing them into those that have a low significance match to SWISS-PROT/PIR and those that are completely unique to the Apicomplexa. For the first set, we compiled a list of assemblies that were similar to at least one other apicomplexan with a p-value < 10–9 based on comparison to SWISS-PROT/PIR and have no similarity outside the Apicomplexa with a p-value < 10–5. Only assemblies with a putative match to SWISS-PROT/PIR of p < 10–9, equal to 20% of all assemblies, were included in this analysis. A total of
105 assemblies were found to be apicomplexan-specific, representing 62
genes of known or putative function. These are summarized in Figure
5, and a complete listing can be found at
http://www.cbil.upenn.edu/paradbs-servlet/pub/SupTable3Api.htm. The
assemblies fall into several broad categories including (1) those that
are specifically restricted to the Apicomplexa; (2) those that are also
known from other organisms, yet are significantly more similar within
the Apicomplexa. Most apicomplexan-specific genes are related to their
unique apical specialization and encode such components as secretory
organelles, surface antigens, and developmental components. Included in
this category are genes encoding secretory proteins such as apical
membrane antigen 1 (AMA-1) that is found in P. falciparum,
T. gondii, and E. tenella. AMA-1 is present on the
parasite cell surface and has been implicated in host cell invasion by
P. falciparum (Peterson et al. 1989; Waters et al. 1990) and
T. gondii (Donahue et al. 2000; Hehl et al. 2000). Also
included are other microneme proteins, likely involved in cell
recognition (Soldati et al. 2001), the major surface antigens (SAGs;
Boothroyd et al. 1997), and a number of proteins characterized from
sporozoites or oocysts. Another highly conserved, yet unique
apicomplexan protein is the small myosin known as TgMyoA in T.
gondii (Hettman et al. 2000) and MyoA in P. falciparum
(Pinder et al. 1998). This motor protein plays an important role in the
unique form of motility of this group of parasites (Sibley et al.
1998). Additionally, a number of metabolic enzymes are found on this
list including a unique apyrase known as nucleoside triphosphate
hydrolase (NPTase), first characterized in T. gondii (Asai et
al. 1995) and also found in N. caninum (Asai et al. 1998). In
further sequencing, a putative homolog of the NTPase has been
identified from S. neurona (D. Howe, unpubl.), indicating that
this enzyme may be more widely contained in the tissue coccidians.
Several conserved enzymes and signaling molecules common to many taxa are also found on the apicomplexan-specific list, for example, hexokinase, 14-3-3 protein, and hypoxanthine phosphoribosyltransferase. Although not restricted to parasites, they are more highly conserved within the Apicomplexa, indicating that they fulfill specific roles that are unique to the members of this phylum. These similarities are not simply caused by these organisms being closer to each other phylogenetically, as many highly conserved genes found in this group do not fit the criteria for being apicomplexan-specific. As such, the apicomplexan-specific features within conserved genes may provide potential targets for development of selective inhibitors. Because of the limited number of known genes in this phylum, the majority of the assemblies do not have significant similarity to the databases. The similarities among these assemblies from different members of this group, however, will provide hints on their importance in apicomplexan-specific physiology and prioritize them for functional characterization. Therefore, we carried out TBLASTX comparisons among all the apicomplexan assemblies and to human–mouse gene sequences. We identified those apicomplexan assemblies that were not similar to known proteins or human–mouse assemblies (p-value < 10–5), but were similar to at least another apicomplexan assembly (p-value < 10–9). This gave 326 assemblies from T. gondii, 173 from N. caninum, 28 from S. neurona, 36 from P. falciparum, and 82 from E. tenella. These assemblies represent unknown genes that are conserved among Apicomplexa but not found in model animal host, human or mouse. A list of these assemblies can be found at http://www.cbil.upenn.edu/paradbs-servlet/pub/SupTable4Api/.
The creation of a relational database to house gene assemblies built
from EST sequences should greatly facilitate analysis of gene function,
expression, and relatedness among the Apicomplexa. The architecture of
ApiESTDB allows it to be periodically updated by inclusion of new data
from ongoing sequencing projects in these and related organisms.
ApiESTDB is also highly complementary to the recently completed genome
sequences for several malaria species (Carlton et al. 2002
Parasite Culture and Library Construction T. gondii cDNA libraries constructed in the type I RH strain or the type II ME49 strain (B7 clone previously referred to as PDS) have been described previously (Ajioka et al. 1998 ). Tachyzoites of the type III
strain VEG, an isolate from an AIDS patient (obtained from Jack
Remington), were grown in human foreskin fibroblast (HFF) cell
monolayers as described previously (Roos et al. 1994). Tachyzoites were
separated from host cells by passage through 3.0-micron membrane
filters (Nucleopore, Inc.) and centrifugation (Howe and Sibley 1995).
mRNAs were isolated from freshly isolated parasites treated with TRIZOL
(Invitrogen, Inc.). cDNAs were synthesized by oligo(dT) priming from
poly(A) mRNA, sized selected, and directionally cloned into a Uni-ZAP
XR vector (Stratagene).
To obtain libraries from the oocyst stage of the life cycle, mRNA was
isolated from partially sprorulated and fully sporulated oocysts that
were produced in cats. Oocysts were purified by sucrose gradient
flotation and chlorox treatment as described (Tilley et al. 1997
N. caninum
S. neurona
E. tenella
P. falciparum
Clone Propagation and DNA Isolation
Then 5 µL of each cDNA glycerol stock in a 96-well microtiter plate
was transferred by a Tango (Robbons Scientific) robot to each well of a
96-well Beckman block, containing 1 mL of Terrific Broth (Difco) with
the appropriate antibiotic. Blocks were incubated at 37°C for 24 h
with agitation at 295 rpm. Cells were pelleted and processed as
described previously (Marra et al. 1999a
DNA Sequencing
Processing and Annotation
The methods have been previously described (Hillier et al. 1996 Because all of the parasites examined here were grown in the presence of host cells, it is possible that some ESTs actually correspond to host mRNAs. To determine the frequency of this we compared the ESTs to sequences in GenBank nr (restricted to the following taxa: Homo, Mus, Xenopus, Gallus, Rattus, Bos, Danio, Sus) using BLASTN and a cutoff of >95% identical over a region of >100 nt. The following percentages of sequences matched this criteria: Toxoplasma (1.0%), Eimeria (0.05%), Neospora (0.6%), Sarcocystis (1.3%), and Plasmodium (0.75). Because it can be difficult to determine true contaminants from highly similar but distinct genes, all ESTs were submitted to GenBank with only a general qualifying statement "that a small proportion of ESTs represent host contaminants and investigators should be aware of this possibility."
Identification of Gene Similarities Prior to Submission
Clustering and Assembly of EST/mRNA Sequences
Annotation of Consensus Sequences
Identification of Phylogenetically Restricted Genes
Redundancy
Stage-Specific Expression
Identification of SNPs
Phylogenetic Analyses
http://blast.wustl.edu; BLAST, Washington University. http://genome.wustl.edu/EST/; EST databases at Washington University. http://PlasmoDB.org/; Plasmodium genome database, University of Pennsylvania. http://prodes.toulouse.inra.fr/prodom/doc/prodom.html; Protein Domain ProDom database. http://www.allgenes.org; All Genes database, Jonathan Crabtree. http://www.biomedcentral.com/1471-2105/3/2; RNA structure database. http://www.cbil.upenn.edu/downloads/DoTS/AllGenesRelease4.0/; AllGenes database, The Computational Biology and Informatics Laboratory, University of Pennsylvania. http://www.cbil.upenn.edu/paradbs-servlet/; ApiESTDB database, University of Pennsylvania. http://www.cbil.upenn.edu/paradbs-servlet/pub/SupTable1Top25.htm; Supplemental data Table 1, University of Pennsylvania. http://www.cbil.upenn.edu/paradbs-servlet/pub/SupTable2Stage.htm; Supplemental data Table 2, University of Pennsylvania. http://www.cbil.upenn.edu/paradbs-servlet/pub/SupTable3Api.htm; Supplemental data Table 3, University of Pennsylvania. http://www.cbil.upenn.edu/paradbs-servlet/pub/SupTable4Api/; Supplemental data Table 4, University of Pennsylvania. http://www.ebi.ac.uk/clustalw/; ClustalW alignment, European Biotechnology Institute. http://www.nature.com/nature/malaria/index.html; Nature publishing Web site for the recent publication of malarial genomes. http://www.ncbi.nlm.nih.gov/cgi-bin/COG/; Conserved orthologous genes, NCBI. http://www.ncbi.nlm.nih.gov/Structure/cdd/cdd.shtml; Conserved domain database, NCBI. http://www.ncbi.nlm.nih.gov/Taxonomy/taxonomyhome.html/; Taxonomic classifications, NCBI. http://www.niaid.nih.gov/dmid/malaria/malrep/; Malaria reagent repository MR4, NIH. http://www.paracel.com/publications/cap4_092200.pdf; CAP4 algorithm, Paracel.
Support for the sequencing was provided by the USDA (NRI 99-35204-8615 to L.D.S.), the NIH (AI45806, AI45806A1, AI48121 to L.D.S.), the Amerman Family Foundation (to D.K.H. at the University of Kentucky), Merck Research Laboratories, and the Burroughs Wellcome Fund (to C.S. and D.S.R.). We thank Jonathan Crabtree for providing tools for Web interface development and Gregory Grant for advice on statistical analyses. We are grateful to members of the Washington University Genome Sequencing Center EST, Processing and Annotation Teams, members of Penn Center for Bioinformatics, and members of the Sibley and Roos laboratories who provided advice and assistance. We also thank Jim Ajioka and Depobam Chakrabarti for providing cDNA libraries, Jane Weismann for expert assistance with data submissions at NCBI, and Michael Gottlieb for his untiring support of parasite genomics. The publication costs of this article were defrayed in part by payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 USC section 1734 solely to indicate this fact.
9 Corresponding author.  E-MAIL sibley{at}borcim.wustl.edu; FAX (314) 362-3203. Article and publication are at http://www.genome.org/cgi/doi/10.1101/gr.693203.
|
Top
ABSTRACT
RESULTS AND DISCUSSION
distribution shape of 0.5. The tree was constructed using
neighbor-joining. The only differences between the trees generated by
these two methods were the relationships among the Plasmodium
species. Otherwise, the trees were identical. Only the neighbor-joining
distance tree is shown. All phylogenetic analyses were performed with
PAUP*4.0 (Swofford 1998