The metabolic reconstruction was carried out with MPMP and KEGG pathways as templates using the genomes and the annotations from EuPathDB. The proteins were annotated to functions by comparing  predicted protein sequences to protein sequences of related species from EuPathDB and KEGG by bidirectional BLASTP and by comparing  the protein signature motifs with InterProScan. Any available biochemical/physiological evidence were included in the analysis. The resources used for metabolic reconstruction of these apicomplexan genomes is illustrated in Figure 1.



Figure 1 - The outline chart details the resources and tools used in general for metabolic reconstruction of the apicomplexan species.


The Figure 2 illustrates the methodology used for the reconstruction of T. gondii genome in detail. The MPMP pathways were used in cases where the metabolic capability of T. gondii is same as or lower than Plasmodium falciparum and KEGG pathways were used where the metabolic capability is either reduced or absent in P. falciparum. A relational database was developed with two tables. One of them contains EC numbers and the corresponding MPMP pathways kindly provided by Prof. Hagai Ginsburg, developer of MPMP metabolic pathways.  The second table contains gene ids and their EC number annotations (if available) of T. gondii ME49 strain retrieved from ToxoDB.  The relational database, bidirectional BLASTP and motif finding tools such as InterProScan were used to annotate genes to enzyme functions within metabolic pathways which were same between P. falciparum and T. gondii. The Toxoplasma non-enzymatic proteins which fit into MPMP pathways are also annotated by carrying out bidirectional BLASTP with P. falciparum orthologues and comparing the protein motifs to the motif library of InterProScan database. The biochemical/physiological evidence available in literature was also used in the pathway reconstruction. We also used the available organellar protein localization evidence (experimental or predicted) to distinguish between potential isoforms of parologous genes in cases when pathway was shown to operate in different organelles. The gene encoded sequences of enzymes belonging to metabolic pathways absent in P. falciparum and present in T. gondii were analyzed by carrying out bidirectional BLASTP to proteins from phylogenetically various organisms which possess the metabolic capability of interest.  For better precision protein motifs of potential candidates were analyzed by means of InterProScan. We also have taken in account any existing biochemical or genomic evidence to confirm our predictions and include the defined functions in the reconstructed pathways.



Figure 2 - Flow chart illustrating the complete process of semi-automatic curation of T. gondii metabolic pathways.


Each metabolic pathways page presents a summary of the importance of metabolic capability and selected biochemical/physiological evidence available in literature at the top of the page. The genes annotated to each function in metabolic pathways are provided in a table with links to respective genes in respective EuPathDB database (e.g ToxoDB for T. gondii) for each gene id. In addition, the EC numbers are linked to the respective enzyme pages of the EC numbers in ExPASy. Any available evidence of protein organelle localisation from different sources were also included in the tables of enzyme annotation for T. gondii. These localisation evidences are not available for other organisms because of lack of experimental data for organisms other than P. falciparum and T. gondii in literature/databases. The source from which the evidence is taken was also included in the table for T. gondii. The list of sources from which localisation data is taken is given in the table below.


Localisation evidence source Description
Apiloc This is a curated resource of localisation of proteins to different organelles. These curations are based also on the presence of experiemental evidence. This source provides localisation annotations for proetins from various apicomplexans species. This resource is also integrated to EupathDB.
Previous publication This publication is a proteomics analysis of Toxoplasma gondii genome from Xia et al. This provides bioinformatics predictions of organellar localisations for Toxoplasma proteins which have proteomics evidence. These predictions were made using bioinformatics prediction tools such as SignalP, TMHMM, PATS, PlasMit and WoLF PSORT. 
Orthology transformation This is only available for apicoplast localisation. The T. gondii orthologues of proteins which are targeted to apicoplast in P. falciparum and present in PlasmoDB P.f subcellular localisation section are provided with this source name.
GO annotation This is the gene ontology cellular component annotation available in the ToxoDB. This is based on JCVI genome annotation or InterPro motif identifications.


The pathways were drawn using CellDesignerTM versions 4.1 and 4.2 and the KEGG links for each metabolite in the pathways were added to the Portable Network Graphics (PNG) image maps. These PMG image maps of pathways were added to the respective pathway page in this website using the iframe plugin. The guidance on understanding shapes of metabolites and proteins and reaction arrows are present in the Guidance page.


The metabolic pathway pages also have a table showing the sources of substrates of the pathways (upstream pathways) and the fates of products synthesised in the pathways (downstream pathways). These data connect the pathways with one another and provide an overview of metabolites which are obtained from host and released to the host. The metabolic pathway names in the table are linked to their original metabolic pathway pages.