Introduction The fungus Nectria haematococca, commonly referred to by its asexual name Fusarium solani, is a member of a monophyletic clade that includes over 50 phylogenetic species known as the “Fusarium solani species complex” ,. Members of the F. solani species complex are able to colonize an impressive variety of environments. As saprobes, they are present in agricultural and non-cultivated habitats, such as forests, scrub communities, savannahs, prairies, swamps, littoral zones, coastal zones, and deserts . As pathogens, members of this complex are responsible for disease on ∼100 genera of plants , and they represent one of the most important group of pathogens associated with opportunistic fungal infections and keratitis in humans , –. Because of their diverse host range, some members have been proposed for the biological control of weeds and other pathogens –. Extreme environments are not beyond the reach of these fungi. F. solani is among the fungal species recovered from the highly radioactive inner parts of the damaged nuclear reactor at Chernobyl . These fungi are capable of growing in anaerobic conditions in the soil  and are tolerant to many compounds shown to be toxic to other fungi ,. F. solani also has been found growing in the caves at Lascaux, France where it is damaging the 15,000 year-old paintings . The ability of these fungi to adapt to so many different environments reflects their genetic plasticity and metabolic diversity. Individual members of this species complex can degrade hydrocarbons, organofluorine compounds, lignin, metal cyanides, and pesticides in the soil –. The most extensively studied species of the F. solani complex is “mating population” (MP) VI of N. haematococca (also called Haematonectria haematococca ). The term “mating population” defines a group of isolates that are sexually fertile with one another, indicating that they are a biological species. Like other F. solani species, N. haematococcaa MPVI isolates can live in many habitats  and classical and molecular genetic analyses have demonstrated that the genes controlling the ability of individual isolates to colonize specific habitats are located on conditionally-dispensable supernumerary chromosomes (“CD chromosomes”), which were first described in fungi in 1991 using N. haematococca MPVI . CD chromosomes are defined as supernumerary chromosomes that are not required for growth under all conditions but confer an adaptive advantage in certain habitats . Subsequent research has demonstrated that in N. haematococca MPVI, genes on these chromosomes are involved in resistance to plant antimicrobials, utilization of specific carbon and nitrogen sources, and in host-specific pathogenicity –. In addition, the properties of the chromosomes and the properties of the genes on these chromosomes, suggest that some of these genes and perhaps even the entire chromosome(s) might have been acquired through horizontal gene transfer (HGT) and have properties similar to the genomic islands of bacteria . The genomic sequence of N. haematococca MPVI has not only the potential to reveal a multitude of metabolic pathways involved in inhabiting many different types of environments, but also to expand on our understanding of the impact of gene flow on fungal evolution. Results General genome features Optical mapping revealed that N. haematococca MPVI isolate 77-13-4 has 17 chromosomes ranging from 530 kb to 6.52 Mb and that the physical size of the genome, 54.43 Mb, is larger than that of any other published ascomycete (Table S1). This is 15% greater than that of the most closely related sequenced fungus, Fusarium graminearum (sexual name: Gibberella zeae), which is known to have undergone significant gene expansion itself . The average gene length (1.67 kb), number of exons (3.08), intron size (84 nt), and the size of the encoded protein (480 aa) (Table S2) are similar to other sequenced ascomycetes ,. Of the gene families in N. haematococca MPVI with at least ten genes, 77% (226 of 293) have more genes than their counterpart in F. graminearum; 18% have more than twice as many genes (Figure S1). As might be expected for a metabolically diverse fungus that can live in so many habitats, among the gene families with the greatest numerical increases are carbohydrate-active enzymes, oxidoreductases, and various monooxygenases and dioxygenases (Table S3). Chromosomal location of genes similar to other fungi To determine why N. haematococca MPVI might have more genes than F. graminearum and to see if these “extra” genes are similar to genes from other fungi, the proteome of N. haematococca MPVI was compared to the genomes of eight other sequenced fungi (F. graminearum, Aspergillus oryzae, A. nidulans, Coccidioides immitis, Chaetomium globosum, Magnaporthe oryzae, Neurospora crassa, and Saccharomyces cerevisiae). The predicted genes of N. haematococca MPVI were classified into three groups: those most similar to F. graminearum, those most similar to genes from the other fungi used in the comparison, or those with no similarity to genes from any of the included genomes. 61.5% were most similar to F. graminearum genes, 28.5% were more similar to the genes from other fungi, and 6.4% had no good match with any of the other genomes. Among the genes with the highest similarity outside F. graminearum, the highest similarity was to Aspergillus species (1786 genes), particularly to A. oryzae (811 genes). The percentage of the genes in each category also was determined for each chromosome. With the exception of chromosome 7, the majority (>60%) of the genes on the large chromosomes (chromosomes 1–10; ranging in size from 6.52 to 3.00 Mb) are highly similar to genes found in F. graminearum (Figure 1), suggesting that these chromosomes are largely derived from an ancestor common to both N. haematococca MPVI and F. graminearum. For chromosomes 7 (3.83 Mb) and 11–13 (2.72–2.19 Mb about half of the genes are more similar to genes of fungal species other than F. graminearum. Interestingly, most of the genes on chromosomes 14–16 (1.57–0.56 Mb) are more similar to genes in other fungi than to genes in F. graminearum, suggesting that these chromosomes are either enriched for very ancient sequences lost from F. graminearum, or the genes were horizontally transferred into N. haematococca MPVI from distantly related fungi. In particular, 20.3%, 19.7%, and 41.3% of the genes on chromosomes 14, 15, and 16, respectively, are most similar to sequences from Aspergillus species. Chromosome 14 also corresponds to a previously studied CD chromosome that carries a cluster of genes for pea pathogenicity (PEP genes) . More than 50% of the proteins encoded by genes on chromosome 17 (530 kb) have no significant similarity to genes from any of the eight fungi selected for comparison (Figure 1) suggesting it is also enriched for very ancient sequences or the genes were derived by horizontal transfer. 10.1371/journal.pgen.1000618.g001 Figure 1 TBLASTN analysis of genes on each chromosome. The relative frequency of the best TBLASTN hits for proteins from each N. haematococca MPVI chromosome. The red line depicts hits to the F. graminearum genome, the yellow line depicts hits to one of the seven other fungal species, and the blue line represents hits to none of the fungal species included in the search. Presence of orthologs, unique genes, duplicated genes, and possible “pseudoparalogs” Many gene families in N. haematococca MPVI are larger than the same families in other ascomycetes. In an effort to investigate further the origin of these additional genes, a phylogenetic analysis was carried out on five gene families (ABC transporters, carbohydrate-active enzymes, P450 monooxygenases, binuclear zinc transcription factors, and chromatin genes; Tables S4, S5, S6, S7, S8). This analysis divided the extra genes into two groups: 1) genes specific to N. haematococca MPVI that are not found in F. graminearum and other fungi, and 2) genes that are present as multiple copies in N. haematococca MPVI but are commonly represented by a single copy in other fungi. In some cases the multiple copies (i.e., paralogs) appear to result from lineage-specific gene duplication (Figure 2). However, in other cases the paralogs are more closely related to a gene from distantly related fungal species (often an Aspergillus species); thus, the gene phylogeny does not reflect the species phylogeny (Figure 2). A specific example of this phenomenon is shown in Figure 3 using the phylogenetically conserved ABC transporter gene YOR1. The YOR1 homologs in 27 fungi were identified by a protein similarity search, and their phylogenetic relationship determined (Figure 3). N. haematococca MPVI has two copies of YOR1 (Nh63546 and Nh73313). Nh63546 appears to be an ortholog of YOR1 in F. graminearum (FGSG_07325), which is the closest relative of N. haematococca MPVI included in this analysis. In contrast, Nh73313 does not demonstrate the expected phylogenetic placement (Figure 3). While grouped with other YOR1 homologs, Nh73313 appears distantly related to the F. graminearum YOR1 and its N. haematococca MPVI ortholog, Nh63546. Genes that demonstrate an incongruent phylogenetic topology, as illustrated in Figure 2 and as specifically shown for Nh73313 in Figure 3, have been called ‘pseudoparalogs’ . A pseudoparalog is a copy of a gene that appears paralogous in a single genome analysis, but when sequences from another genome are included, it appears as if the gene were transferred laterally into the genome. However, it has been pointed out recently that the same topology can occur if there is gene duplication, diversification, and differential gene loss (DDL) ,. Specific and duplicated genes were observed within all five gene families and apparent pseudoparalogous genes were found in all families except the chromatin genes. An example of a pseudoparalog for each gene family is given in the footnotes of Tables S4, S5, S6, S7. 10.1371/journal.pgen.1000618.g002 Figure 2 Phylogenetic placement of paralogs in N. haematococca MPVI. N. haematococca 1 is the ortholog. (A) Placement of a gene at this position implies a recent gene duplication. (B) Placement of a gene at this position indicates the gene may be a pseudoparalog. 10.1371/journal.pgen.1000618.g003 Figure 3 The phylogenetic relationship of the ABC transporter YOR1 from selected fungal genomes. Maximum parsimony analysis was used to establish the phylogenetic relationship between the ortholog (Nh63546, red box) and the pseudoparalog (Nh73313, blue box) of N. haematococca MPVI. Since it has been proposed that HGT could account for some of the genes on the 1.6-Mb CD chromosome of N. haematococca MPVI ,, and four of the five expanded gene families included pseudoparalogs, a global analysis of the genome was undertaken to identify possible pseudoparalogs and genes unique to N. haematococca MPVI. Reciprocal BLASTp searches between the F. graminearum and N. haematococca MPVI proteomes resulted in the identification of 8,922 possible orthologs representing 56.8% of the genes in N. haematococca MPVI. The remaining 6,785 genes in N. haematococca MPVI were identified as ‘unique’ genes. It is within these unique genes that pseudoparalogs are found. To identify possible pseudoparalogs, the unique genes from N. haematococca were compared to the F. graminearum proteome and the orthologs of N. haematococca MPVI with a reciprocal BLASTp approach. A liberal arbitrary cut off of 40% identity over a 40-amino acid length was used to limit the results. A non-stringent cut off for orthologs was used as it created a more comprehensive search for possible pseudoparalogs. Those unique genes that had mutual best hits to both genes of a F. graminearum-N. haematococca MPVI ortholog-pair were classified as possible pseudoparalogs. For example, two CAX (calcium exchange) transporter genes were found in this set; one (Nh65123) is orthologous to F. graminearum FGSG_01606 and the phylogenetic placement of the second, Nh101770, suggests it is a pseudoparalog (Figure S2). Using this approach, 1,331 possible pseudoparalogs were identified (Figure 4). It should be noted that this approach does not differentiate between duplicated and pseudoparalogous genes. 10.1371/journal.pgen.1000618.g004 Figure 4 Chromosomal locations of possible pseudoparalogs. The percentage for each chromosome is based on the number of possible pseudoparalogs out of the total number of genes on that chromosome. The G+C percentage and codon usage of orthologs, unique and possible pseudoparalogs in the N. haematococca MPVI genome Outside of the A+T rich repeated regions typically associated with pericentromeric or centromeric regions, the G+C content is generally consistent among genes within a genome ,. However, sequences introduced into a genome sometimes retain characteristics of the donor genome. This observation has led to the use of G+C content and codon usage to identify regions in prokaryotic genomes that might have arisen via HGT ,,. The large data set of the groups of genes found in N. haematococca MPVI allowed an analysis of the G+C content of the orthologs to F. graminearum, N. haematococca MPVI unique genes, and possible pseudoparalogs. The overall %G+C content of the orthologs was 55.2% versus 53.3% for the unique genes (P = 2 kb), the cloning vector (pMCL200cDNA), and the sequencing primers (Fw: 5′-AGGAAACAGCTATGACCA-3′, Rv: 5′-GTTTTCCCAGTCACGACGTTGTA-3′) . 24,793 ESTs were obtained from mycelium grown in the PDB medium and 8,327 from the mycelium treated with pisatin. Genome finishing methods Initial read layouts from the whole genome shotgun assembly were converted into a Phred/Phrap/Consed pipeline  and, following manual inspection of the assembled sequences, finishing was performed by resequencing plasmid subclones and by walking on plasmid subclones or fosmids using custom primers. All finishing reactions were performed with 4∶1 BigDye to dGTP BigDye terminator chemistry (Applied Biosystems). Repeats in the sequence were resolved by transposon-hopping 8-kb plasmid clones. Fosmid clones were shotgun sequenced and finished to fill large gaps, resolve large repeats, and to extend into chromosome telomere regions where possible. After finishing, the genome remained in 209 scaffolds as a result of many regions of the genome being apparently unclonable in the shotgun libraries constructed for this project. The resulting assembly was joined and validated by alignments to a N. haematococca optical map (generated by digestion with NheI), with 92.26% of the sequence (72 scaffolds or 47,191,137 bp) being placed onto 17 chromosome optical maps. 3,958,438 bp of the 137 smaller scaffolds remain unplaced because of the lack of sufficient restriction sites. The genome consists of 51,149,575 base pairs of finished sequence with an estimated error rate of less than 1 error in 100,000 bp. Optical mapping Protoplasts of 77-13-4, made as described above , were directly lysed with a protoplast lysing solution (10 mM EDTA, 5 mM EGTA, 1 mg/ml proteinase K, pH 8.0) by heating to 65°C for 30 min to 1 hr and then incubating overnight at 37°C. The protoplast concentration was adjusted to ∼700 protoplasts per microliter. Optical mapping operations followed previously published techniques ; briefly, randomly sheared high molecular weight DNA was loaded onto optical mapping chips for restriction digestion by NheI (New England Biolabs). DNA was stained with YOYO-1 fluorochrome (Invitrogen) and the chips were scanned on an automated fluorescence microscope system for image capture, analysis, and map construction . Resulting single-molecule restriction maps were assembled into genome-wide contigs , that served as map scaffolds for sequence joining and validation efforts. Gene prediction and automated annotation Gene models (15,707) were predicted and automatically annotated using the Joint Genome Institute (JGI) Annotation Pipeline. Several gene predictors were used on repeat masked assembly: ab initio FGENESH and homology-based FGENESH+  and Genewise . The predicted gene models were verified, corrected, and extended using 33,142 N. haematococca MPVI ESTs. All predicted gene models were functionally annotated by homology to proteins from the NCBI non-redundant set and classified according to Gene Ontology , eukaryotic orthologous groups (KOGs ), and KEGG metabolic pathways . Of the 15,707 models, 93% were complete models, 25% were supported with EST aligment, 94% with NR alignment, 73% with Swissprot alignment and 52% with Pfam alignments. For every locus the ‘best’representative model was selected based on EST and homology support, to produce a non-redundant representative set, subject to manual curation and the analysis described here. Best fit analysis of N. haematococca to other fungi Each predicted protein from the N. haematococca MPVI genome was used as a query in a TBLASTN search of a database consisting of the genomes of A. oryzae, A. nidulans, C. globosum, C. immitis, F. graminearum, M. oryzae, N. crassa, and S. cerevisiae. For each protein query, the genome with the best hit below a threshold of 1e-5 was identified. Phylogenetic analysis The predicted amino acid sequences for hypothetical proteins were aligned with ClustalW 1.81 . The resulting alignment files were imported into MacClade 4.08  for manual editing and exclusion of all ambiguously aligned regions. Heuristic searches for maximum parsimony (MP) were conducted in PAUP* , and neighbor-joining distance trees were generated in MacVector 10.0.2 (Symantec Corporation). Statistical support was calculated from 1,000 bootstrap replicates. RIP index To identify repetitive regions of the genome in the absence of a curated repeat library, the genome was divided into 1 kb non-overlapping windows and BLASTN was used to align each against the complete genome. If the only match greater than 500 bp was a self-match, then the window was labeled as unique, otherwise it was labeled as repetitive. For each 1 kb window, the ratio of TpA/ApT frequency was calculated. Experimental demonstration of RIP A transformant (Tr78.2) of N. haematococca that contained several copies of the hph gene from the transformation vector pCWHyg1 , was crossed (cross 370) with 77-15-7 as previously described . The hph gene is linked to the homoserine utilization phenotype (HUT) in Tr78.2 and 370 progeny were screened for hygromycin sensitivity and HUT. All forty progeny from cross 370 were sensitive to hygromycin, and half were HUT+. DNA was isolated from two hygromycin sensitive/HUT+ isolates (370-4 and 370-8) as previously described . Sequences of hph were amplified using PCR and the primers Hyg-F (5′-CGGAGATTCGTCGTTCTGAAGAG-3′) and Hyg-R (5′-TTCTACACAGCCATCGGTCCAG-3′) following the manufacturer's protocol (Invitrogen) and the following set of conditions (94°C for 45 sec, 59°C for 30 sec, 72°C for 1.5 min, for 35 cycles). The resulting 1,242 bp products containing hph were cloned into the pGEM T-EZ vector (Promega Corporation, Madison, WI) and hph was sequenced using the Hyg2-F (5′-ACGCGACAACTGAGTGACTG-3′) primer adjacent to hph. Sequencing was done by the Genomic Analysis and Technology Core (GATC) facility at the University of Arizona. Pulsed-field gel electrophoresis (PFGE) analyses of chromosome-sized DNA 77-13-7 and its derivative B-33 have been described , as have 77-13-4 and its derivative HT-1 . The preparation of chromosome-sized DNAs and the PFGE were performed essentially as described previously ,. For making protoplasts of 77-13-7 and B-33, the enzyme mixture of Garmaroodi and Taga  was used, while the enzyme mixture of Rodriguez-Carres et al.  was used for 77-13-4 and HT-1. Protoplasts (ca. 3×108 protoplasts/ml) were embedded in low melting temperature agarose (Bio-Rad Laboratories Inc., Hercules, CA) and chromosome-sized DNAs were separated in 0.5× TBE buffer on a 1% (w/v) pulsed field certified agarose (Bio-Rad Laboratories) gel with a contour-clamped homogeneous electric field apparatus (CHEF-DR II, Bio-Rad Laboratories). Running conditions were 5.4 V/cm and pulse time of 120 s for 13 h followed by 180 s for 13 h. Chromosomal DNAs of S. cerevisiae (Bio-Rad Laboratories) were used as the size markers. Attempts to detect chromosome-specific sequences found on the ends of supernumerary chromosomes Genomic DNA of N. haematococca MPVI was isolated as previously described . PCR was used to test for the presence of sequences found on the ends of the assembled sequences representing chromosomes 14, 15, and 17 in isolates 77-13-7, 77-13-4, B-33, and HT-1. Primers were designed to amplify regions of scaffolds located on the ends of the respective chromosomes: scaffolds 24 and 62 for chromosome 14, scaffolds 25 and 57 for chromosome 15, and scaffolds 74 and 90 for chromosome 17. These sequences were blasted against the N. haematococca MPVI genome sequence to confirm that they were not found on chromosomes other than those of interest. Sequences of the primers (Invitrogen) used for PCR were as follows: for chromosome 14, scaffold 24: forward primer, 5′-GCCAGGAGATCGAGATATGA-3′ and reverse primer, 5′-GTGGATGAGATCGGTGTTTC-3′; for chromosome 14, scaffold 62: forward primer, 5′-CTCCATCTTCTCGGCAATGT-3′ and reverse primer, 5′-CTTGGTTCACTCGCATACTTG-3′; for chromosome 15, scaffold 25: forward primer, 5′-GACCGTCAAGGGAGCTACAG-3′ and reverse primer, 5′-ATCAGGGGTCATGTGAAGC-3′; for chromosome 15, scaffold 57: forward primer, 5′-GGCCTTTGTACTCGCATTTA-3′ and reverse primer, 5′-GACCCTCTGCCTTCTTCTTC-3′; for chromosome 17, scaffold 74: forward primer, 5′- CGCCCACTTCTTTGTCTCTA-3′ and reverse primer, 5′-AGCGAATTCATTTGAAGCAG -3′; and for chromosome 17, scaffold 90: forward primer, 5′-GGAGACGTTGATGAGATTGG -3′ and reverse primer, 5′-CATCTGTTGAACCCACACAA -3′. Each reaction had a total volume of 50 µl containing 300 nM forward primer, 300 nM reverse primer, 1 µl (∼50 ng) DNA template, 5 mM MgCl2, 5 µl 10× PCR buffer, 200 µM (each) of dATP, dTTP, dCTP, and dGTP, and 1 unit Taq DNA Polymerase. The PCR protocol consisted of an initial denaturation step of 95°C for 3 min., 35 cycles at 95°C for 30 sec, 55°C for 30 sec, and 72°C for 30 sec, and a final elongation step at 72°C for 30 sec. PCR products were run on a 0.8% agarose gel containing ethidium bromide and visualized under UV light. Genomic DNA of N. haematococca isolates 77-13-4 and 77-13-7 was used as positive controls. Search for segmental duplications BlastP searches of the protein set against itself with a threshold value of 1e-20 identified 2,259 gene pairs as best-bidirectional hits. Segmental duplicated regions were defined as genomic regions that share at least three genes in the same order and orientation, while the distance between neighboring gene pairs is less than 50 kb. Supporting Information Figure S1 Ratio of the number of genes in gene families in N. haematococca MPVI versus F. graminearum. Only gene families that had ≥10 members in N. haematococca MPVI were used in the analysis. The number of genes per family was derived from Interpro calls made by the JGI for N. haematococca MPVI, and by the Munich Institute for Protein Sequences (MIPS) for F. graminearum. (1.27 MB TIF) Click here for additional data file. Figure S2 The CAX (calcium exchanger) transporter clade from select fungal genomes. Maximum parsimony analysis was used to establish the phylogenetic relationship between the ortholog (Nh65123, red box) and the pseudoparalog (Nh101770, blue box) of N. haematococca MPVI. (8.06 MB TIF) Click here for additional data file. Figure S3 Distibution of repeat identity. NC7 is N. crassa, MG5 is M. oryzae, AN1 is A. nidulans, “FG3 Repeats” is F. graminearum and “FS Repeats” is N. haematococca MPVI. (4.09 MB TIF) Click here for additional data file. Figure S4 Effect of RIP on a family of telomere-associated helicases (TAH) in N. haematococca MPVI. Partial alignment of the 12 predicted TAH genes (tah), spanning only the first three conserved motifs. The top row shows the predicted translation of the fourth tah gene on scaffold 45 (TAH_45-4). While many mutations occur in the wobble position, note the presence of nonsense codons (*). Nucleotides in red (G to A change) and orange (C to T change) can be explained by a single RIP-type mutation, while nucleotides in pink denoted non-RIP-type transversions. Conversion of RIP-type C∶G to T∶A mutations back to the likely original sequence (“de-RIP”, blue), results in a consensus sequence (TAH_ORI) that closely resembles that of the Metarhizium anisopliae TAH1 sequence (MaTAH1; note absence of nonsense codons in the derived consensus sequence, residues in red indicate changes compared to the TAH_45-4 sequence). De-RIP of the complete coding region results in a single large ORF without predicted introns or nonsense codons, similar to the M. anisopliae TAH1 gene (Inglis PW, Rigden DJ, Mello LV, Louis EJ, Valadares-Inglis MC 2005 Monomorphic subtelomeric DNA in the filamentous fungus, Metarhizium anisopliae, contains a RecQ helicase-like gene. Mol Genet Genomics 274: 79–90). (7.54 MB TIF) Click here for additional data file. Figure S5 Repeat-induced point mutation (RIP) in N. haematococca MPVI. The hygromycin resistance (hph) gene is mutated from G to A at multiple TpG positions (indicated in red) in isolates 370-4 and 370-8. (1.36 MB TIF) Click here for additional data file. Table S1 Comparison of genome statistics of several filamentous ascomycete fungi. (0.06 MB DOC) Click here for additional data file. Table S2 Properties of the genes of N. haematococca MPVI. (0.03 MB DOC) Click here for additional data file. Table S3 Gene families that are at least two-fold larger in Nectria haematococca MPVI than in Fusarium graminearum. (0.09 MB DOC) Click here for additional data file. Table S4 The number of ABC transporters in Nectria haematococca MPVI compared to other fungi. (0.05 MB DOC) Click here for additional data file. Table S5 Carbohydrate-active enzymes in N. haematococca MPVI compared to other fungi. (0.07 MB DOC) Click here for additional data file. Table S6 The number of cytochrome P450 genes in Nectria haematococca MPVI compared to other fungi. (0.06 MB DOC) Click here for additional data file. Table S7 Number of predicted genes in Nectria haematococca MPVI that contain transcription factor motifs compared to other fungi. (0.10 MB DOC) Click here for additional data file. Table S8 The number of chromatin genes in N. haematococca MPVI compared to other fungi. (0.05 MB DOC) Click here for additional data file. Table S9 Distribution of repeat elements in the genome of Nectria haematococca MPVI. (0.08 MB DOC) Click here for additional data file. Table S10 Properties of the chromosomes and genes on each chromosome in N. haematococca MPVI. (0.09 MB DOC) Click here for additional data file. Table S11 The protein kinases of N. haematococca MPVI compared to S. cerevisiae. (0.03 MB DOC) Click here for additional data file. Table S12 The number of polyketide synthases (PKS) and nonribosomal peptide synthetases (NRPS) of Nectria haematococca MPVI compared to other fungi. (0.06 MB DOC) Click here for additional data file. Table S13 Distribution of Small Secreted Proteins (SSP) among filamentous Ascomycetes as identified by SignalP. (0.05 MB DOC) Click here for additional data file.