Review Reports

Round 1

Reviewer 1 Report

The article has detailed data and no major issues, but why are the strain numbers following the Latin names in the main text and Table 1 inconsistent with those in Figure 3B?

The article has detailed data and no major issues, but why are the strain numbers following the Latin names in the main text and Table 1 inconsistent with those in Figure 3B?

Author Response

Please see the attachment

Author Response File: Author Response.pdf

Reviewer 2 Report

The authors analyzed the genome data of four Diaporthe fungus from medicinal plants to extract the biosynthetic gene clusters and the gene cluster families using bioinformatic tools. As a result, conserved and isolated GCFs were found across the four fungi, thus unveiling the biosynthetic potentials of these fungi.

1. The introduction does not give sufficient background to address the knowledge gap from the literature. The authors need to tell the readers why the authors interested in these four strains of Diaporthe. Also, any more information about the medicinal plants? 
2. Would that be any publicly available genome of Diaporthe? It should be yes, at least there is an example, as illustrated in lines 77-81, also ref. 14. Then, what's the differences in genome between the referenced strains and the four titled Diaporthe? In other words, those known and newly reported fungi (i.e, the four reported in this study) should be compared in a larger context using the same methods as describe in the method section to find out the biosynthetic insights.
3. The knowledge of the secondary metabolites of Diaporthe should be fully described and discussed in the text, where appropriate, to connect the predicted BGCs to the reported metabolites of this genus. In this manuscript, I did not see such analysis, or only at a minimum content that I did not notice. 
4. Meanwhile, some BGCs are predicted to produce known metabolites, and have these compounds been reported in Diaporthe as well? Please fully discussed.
5. As mentioned by the authors, some Diaporthe Members "... occur as endophytes, saprobes, and pathogens, illustrating the ecological versatility of Diaporthe–plant interactions (lines 56-57)". Are there some specific BGCs found only in endophytes when compare to those saprobes, and pathogens? This should be interesting to the readers.

Minor points:

1. In the method and material section, section 2.1, lines 116-121, What's the accession number for the genome of each fungus? These data should be public available for the purpose of knowledge sharing.
2. The references section should revise according to the MDPI/J of Fungi requirements. For instance, the journal names should list as abbreviation, so the references, such as 10, 11, 15, 46, should be revised.

Author Response

Please see the attachment

Author Response File: Author Response.pdf

Reviewer 3 Report

The manuscript presents a comparative genomic analysis of four medicinal plant-associated endophytic Diaporthe isolates, integrating orthology analysis, functional annotation, and genome mining using antiSMASH and BiG-SLiCE to characterize their biosynthetic potential. The study is well conceived and addresses a relevant topic in fungal comparative genomics and natural-product discovery. 

However, the following issues needs to be addressed before the manuscript can be considered for publication. 

1.  The manuscript states that 15 of the 348 GCFs were “conserved across all four isolates.” However, the Methods define a core GCF as one in which at least one BGC member occurs in each of the four isolates. Thus, this terminology describes shared GCF membership rather than demonstrating that the corresponding BGCs are identical, highly similar across their entire length, or strictly orthologous. This distinction is important because the current wording could lead readers to interpret the 15 GCFs as representing highly conserved BGCs across the four genomes. I recommend that the authors use terms such as “core GCF,” “shared GCF,” or “GCF represented in all four isolates” where appropriate and explicitly distinguish GCF-level conservation from sequence- or synteny-level conservation.

2. Section 2.4 states that BGCs lacking MIBiG matches above a 45% threshold were designated “chemically uncharacterized.” In Section 3.6, however, the terminology is applied at the GCF level, and some GCFs are described as containing a characterized-cluster match in only one or a subset of their members.The authors should explicitly define the rule used to classify an entire GCF as characterized or uncharacterized.

3. The Methods describe a “cluster-level similarity threshold of 45%” for retaining KnownClusterBlast/MIBiG matches. The authors should define precisely what this percentage represents and how it was calculated or extracted from the antiSMASH results. Because this threshold directly determines which BGCs are treated as characterized pathway-associated clusters, the metric and rationale should be clearly documented. Please also report the relevant antiSMASH/MIBiG database version used for the analysis and provide sufficient methodological detail for the identification and filtering of MIBiG-associated clusters to allow the analysis to be reproduced.

4. The Discussion, particularly Section 4.3, devotes considerable attention to compounds such as squalestatin, nectriapyrone, cercosporin, and related metabolites and their reported biological activities. These associations are potentially useful for prioritizing candidates, but the evidence presented in this study is genomic rather than experimental. The manuscript appropriately acknowledges that genomic similarity should be interpreted as evidence of related biosynthetic capacity rather than confirmation of metabolite production. However, the subsequent discussion sometimes gives greater emphasis to the known biological activities of the reference compounds than to the actual evidence provided by the present study. The Discussion should therefore be reframed so that the primary focus is on the predicted biosynthetic potential of the Diaporthe isolates.

5. The manuscript would benefit from a more critical discussion of the limitations inherent in predicting chemical diversity from genome sequence alone. The presence of a BGC does not establish pathway expression, metabolite production, pathway completeness, or biological activity. Likewise, similarity to a characterized MIBiG cluster does not necessarily imply production of the same compound. These limitations are particularly relevant here because the central conclusions rely heavily on predicted BGCs, GCF assignments, and similarity to characterized pathways. The authors should explicitly discuss possible effects of gene loss, gene duplication, tailoring-enzyme divergence, regulatory differences, incomplete assemblies, and inaccurate gene models on the interpretation of BGC comparisons.

6. The study is entirely computational, which is reasonable for a comparative genomics study, but this limitation should be more clearly integrated into the conclusions and Discussion. The authors should distinguish between identifying candidate biosynthetic loci and demonstrating their functional activity. The manuscript would also benefit from a more concrete discussion of how the identified candidates could be prioritized for experimental follow-up. For example, candidates could be discussed in terms of their degree of genomic conservation, divergence from known pathways, distinctive gene content, unusual core-biosynthetic architecture, or isolate restriction. This would give the natural-product-discovery implications of the study a clearer direction.

7. The Introduction identifies four medicinal plant-associated Diaporthe isolates from three plant species/host contexts, including two isolates associated with Vochysia divergens. The authors should provide a clearer rationale for selecting these particular isolates for comparison.

8. Please clarify whether the isolates were selected because of their phylogenetic relatedness, ecological association, availability of genome assemblies, previously reported metabolite production, or another criterion. It would also be useful to explain whether the four isolates should be considered representative of a particular ecological or phylogenetic group, or whether they were selected primarily as available comparative genomes.

9. The Materials and Methods provide software versions but omit several parameters, and database/version information that are necessary for reproducibility. For example, For OrthoVenn3, BiG-SLiCE, and antiSMASH.

1. The Introduction is generally informative, but the transition from the general biology of fungal endophytes to the comparative genomics of Diaporthe could be more focused. The rationale for the specific four-isolate comparison should be made clearer.

2. The Introduction would benefit from more specific citations when making broad statements about cryptic biosynthetic pathways and the ability of fungal genome mining to reveal biosynthetic potential that is not apparent from metabolite profiling.

3. The manuscript should clearly state that the BGC-class counts in Table 2 are not mutually exclusive and therefore do not sum to the total number of BGC regions. This is especially important because multiclass/hybrid regions can contribute to more than one biosynthetic class. A note in Table 2 would help prevent misinterpretation.

4. The authors provides a detailed inventory of BGC classes, but the biological or comparative significance of the differences among isolates is not always clear. A concise summary highlighting the principal differences among the four repertoires before presenting the detailed class-by-class information would improve readability.

5. The finding that five core GCFs show similarity to the same characterized pathway across all four genomes is potentially important. The supporting evidence should be presented more explicitly, preferably in a supplementary table containing the relevant BGC identifiers, GCF assignments, MIBiG accession numbers, similarity scores, and, where available, information on core-gene or gene-content similarity.

6. Also, the authors should ensure that all figure legends clearly distinguish BGC/GCF membership from direct sequence or functional conservation.

7. The isolate-specific pathway section is informative but reads somewhat like a catalogue of MIBiG matches. Consider providing a short comparative summary at the beginning of the section identifying the main patterns among isolates, followed by the detailed examples.

8. The Discussion is long and frequently repeats observations already presented in the Results. It could be made more concise by reducing repetition and placing greater emphasis on interpretation rather than restatement.

9. Table 1 should be reformatted for readability and consistent alignment.

10. There are numerous typographical and grammatical errors throughout the manuscript. A thorough language edit is needed before publication. For example, the authors should perform a final, manuscript-wide check of species nomenclature and ensure identical spelling and italicization in the main text, tables, figures, and supplementary material.

Author Response

Please see the attachment

Author Response File: Author Response.pdf

Reviewer 4 Report

The manuscript addresses a highly relevant and timely topic concerning the genome mining of biosynthetic gene clusters (BGCs) in micromycetes, focusing specifically on medicinal plant-associated endophytic Diaporthe. The study is methodologically sound and holds substantial value for the field of natural product discovery. However, a number of critical concerns regarding the genomic context and strain selection rationale must be thoroughly addressed before the manuscript can be considered suitable for publication.

 

Major comments

1) Genome Data Availability and Context:

The primary issue relates to the presentation and underlying data of the genome sequences. The authors need to clarify the status and accessibility of the genomic sequences, particularly for three of the four studied fungal isolates. If these are newly sequenced or re-annotated genomes, explicit details regarding repository accession numbers and data availability must be provided.

2) Phylogenetic and Evolutionary Analysis: To provide a robust foundation for the comparative genomics section, a more rigorous taxonomic comparison is required. The authors should construct a comprehensive phylogenetic tree utilizing a broader set of closely related reference strains from the genus Diaporthe. Without this evolutionary framework, the comparative data on predicted orthologous proteins and BGCs lack a clear biological anchor and clear interpretative meaning.

3) Justification of the Study Design and Sample Size: The rationale behind selecting these specific four isolates requires a strong scientific justification. The authors must explicitly state whether this choice was driven by a deliberate conceptual framework (e.g., specific host plant traits, phenotypic differences, or ecological niches) or if it was simply opportunistic due to the immediate availability of public sequence data.

4) Contextualization within the Genus Diaporthe: In modern bioinformatics, comparative pipelines routinely analyze dozens, hundreds, or even thousands of genomes within a given taxonomic group to map out BGC and gene cluster family (GCF) diversity. Against this backdrop, focusing on only four isolates appears narrow. The authors should contextualize their study by addressing the following:

- How many Diaporthe genomes are currently available in public repositories?

- What makes these specific four genomes outstanding or representative?

- Is the genus Diaporthe generally underrepresented in genomic databases, or do these four isolates possess unique, remarkable genomic features that justify this limited sample size?

1) Lines 713–717: The authors state that "The scale of the BGC repertoires observed here is also noteworthy. The four isolates encoded 112–129 BGCs each, indicating considerably greater biosynthetic potential than can normally be inferred from metabolite profiles obtained under a single cultivation condition. A recent genome-mining study of D. kyushuensis ZMU-48-1 identified 98 BGCs..."This statement raises an important question regarding the selection criteria for the studied microorganisms: why were these specific four strains chosen for bioinformatic analysis, especially given their remarkably large BGC repertoires? Was this selection guided by a specific hypothesis, or was it largely random/opportunistic? To strengthen the manuscript, the authors should clarify the broader context in the Introduction (or another appropriate section) by providing the total number of currently known Diaporthe species and explicitly justifying why, out of all available taxa, a comparative genomic analysis was performed exclusively on these four specific isolates.

 

2) Lines 116–121 (Materials and Methods) and Data Availability: The authors state that "Publicly available genome assemblies, structural annotations, and predicted protein sequences were retrieved for each isolate...". However, neither Table 1 nor Supplementary Table S1 contains the necessary repository accession numbers to locate these datasets.

Table 1 outlines general assembly metrics (such as genome size, scaffold count, N50, etc.) but omits public database identifiers.

Table S1 is presented in a raw GFF format (typically generated by gene prediction tools like AUGUSTUS). While valuable for programmatic use, a GFF file is not meant for manual review and is functionally unusable without access to the underlying scaffold or contig nucleotide sequences (.fasta).

For Diaporthe sp. HANT25, a draft genome assembly is accessible in GenBank under the accession number GCA_013435955.1. This specific identifier must be explicitly cited in the text and tables.

For the remaining three isolates (D. inconspicua LGMF1612, D. cf. heveae LGMF1633, and D. vochysiae LGMF1583), public assemblies could not be verified from the current text. To adhere to standard journal data-sharing policies and ensure the reproducibility of the reported BGC profiles, the authors must deposit these raw assemblies into a public repository (e.g., NCBI GenBank, ENA, or DDBJ) and provide the respective accession numbers directly within the manuscript.

 

3) Phylogenetic Context for Comparative Analyses (Lines 694–695 and 737–740):
The authors state that "only 15 of the 348 biosynthetic gene cluster families (GCFs) were conserved across all genomes" (Lines 694–695). Later, in Section 4.2, they note that "The two isolates obtained from Vochysia divergens, D. cf. heveae LGMF1633 and D. vochysiae LGMF1583, also retained distinct BGC repertoires despite their common host association, emphasizing that biosynthetic capacity cannot be inferred from host identity alone"(Lines 737–740).

These observations immediately raise a critical evolutionary question: what is the actual phylogenetic distance between these four genomes? To make the comparative proteomic and BGC analysis biologically meaningful, the authors must reconstruct a robust phylogenetic tree for the genus Diaporthe and closely related outgroup taxa, explicitly including the four studied isolates. Providing this evolutionary framework is essential to assess the true genetic relatedness of the compared organisms and to understand whether the observed divergence in BGC architecture is driven by phylogenetic distance or ecological adaptation.

 

4) Conceptual Redundancy (Lines 51–54 and Lines 71–74):
There is a clear conceptual overlap between two passages in the Introduction.

In Lines 51–54, the authors write: "Advances in genomics and metabolomics have further demonstrated that the chemical potential of fungal endophytes is considerably greater than that observed under routine laboratory cultivation, because many biosynthetic pathways remain silent or weakly expressed under standard conditions [4,5]."

In Lines 71–74, a highly similar sentiment is repeated: "Genome sequencing has revealed that fungal endophytes often possess substantially more BGCs than would be predicted from metabolites detected under a limited set of culture conditions, exposing a large reservoir of cryptic biosynthetic potential [11–13]."

This duplication of the same core idea is redundant. The authors should consolidate these statements, keeping only one of the variants (or merging them seamlessly) to ensure a more concise and precise narrative in the Introduction.

5) Section 2.3 (Functional annotation of predicted proteins):
The authors describe the functional annotation and GO/Pfam domain profiling of the predicted proteomes. To ensure transparency and provide full access to the bioinformatic outputs, the complete underlying datasets for the functional annotations of all four Diaporthe isolates should be provided as a comprehensive table in the Supplementary Materials.

6) Figure 2 (Image Quality and Resolution):
The current resolution of Figure 2 is too low, making the embedded text and charts difficult to read. To improve readability and ensure that readers can effortlessly interpret the presented data, the authors should separate the composite figure parts (2A and 2B) into two distinct, high-resolution figures.

7) Subsection 3.3 Title:

The current title, "Extensive biosynthetic gene cluster repertoires dominated by polyketide pathways," reads more like an informative, narrative sentence than a structured academic subheading meant to generalize the section's contents. Reword the title of this subsection.

8) Figure 5 (Gene Cluster Maps and Visualization Anomalies):
The authors use directional arrows to indicate the transcriptional orientation of genes within the predicted BGCs. However, there are several visual inconsistencies and formatting anomalies that could confuse the reader and must be corrected:

In the first/topmost BGC map (GCF87, Armillyl orsellinate...), the arrow for the 9th gene incorrectly points in both directions simultaneously (a bi-directional arrow). The authors should clarify the biological meaning of this notation or fix it if it is a rendering error.

In the same topmost cluster, the box/arrow height for the 7th gene is noticeably smaller than all other genes in the series. Fungal gene maps should maintain consistent lane and feature dimensions unless a specific box height code is explicitly defined in the legend.

Regarding gene spacing and consistency, if the reduced height or overlapping format was intended to indicate tight intergenic spacing or overlapping open reading frames (ORFs), this logic is not applied uniformly. For instance, in the 5th BGC from the bottom (GCF292, Chaetolivacine A–C), genes 21 and 22 are clustered just as closely but maintain standard, uniform box heights.

To avoid misinterpretation and improve data presentation, the authors must revise the formatting of Figure 5 to ensure complete graphical consistency across all illustrated genomic scaffolds

 

 

Author Response

Please see the attachment

Author Response File: Author Response.pdf

Round 2

Reviewer 2 Report

The authors have addressed all my concerns.

No other comments.

Author Response

Thank you for your valuable suggestions and accepting our manuscript

Reviewer 3 Report

This manuscript presents a comparative genomic analysis of four medicinal plant-associated endophytic Diaporthe isolates. The central conclusion—that these isolates share a largely conserved proteomic background but possess more variable and partly isolate-specific biosynthetic repertoires—is reasonable and potentially useful for natural-product discovery. However, the manuscript currently contains numerical inconsistencies, methodological ambiguities, and nomenclature errors that affect confidence in the central results. The study is computational and hypothesis-generating; this is acceptable if framed clearly and if predicted biosynthetic potential is distinguished from experimentally validated metabolite production.

• The Discussion still devotes considerable attention to the known biological activities of squalestatin, nectriapyrone, cercosporin, and related compounds. These are useful as prioritization clues, but the primary emphasis should be on the predicted biosynthetic potential of the four Diaporthe isolates.

• The authors have appropriately added limitations, but the Conclusion should explicitly state that the identified BGCs and GCFs are candidate biosynthetic loci rather than experimentally validated pathways. The prioritization framework for future experimental follow-up should be concise and concrete.

• The response letter states that the revised classification identified 91 characterized pathway-associated GCFs, 15 partially characterized GCFs, and 242 chemically uncharacterized GCFs, but these counts must be clearly reconciled with the main text, figures, and supplementary tables.

• The Methods state an operational cutoff of ≥45% reference-cluster gene coverage. But Figure 5B legend defines high confidence as ≥70, medium as 30–70, and low as <30.

• Correct nomenclature errors, e.g., Hydrocarpa anthelminticus vs. Hydnocarpus anthelminticus and Diapothe instead of Diaporthe.
• Provide additional reproducibility details in the Methods

  • Phylogenetics: specify the substitution model, rate variation, gap handling, alignment trimming, and partitioning strategy.

  • OrthoVenn3: confirm version and all non-default parameters.

  • eggNOG-mapper: the manuscript says version 3.0, but the cited paper is for eggNOG-mapper v2. Check and correct.

  • antiSMASH: confirm the exact version and settings.

  • BiG-SLiCE: justify the 0.4 distance threshold and consider whether results are sensitive to alternative thresholds.

  • MIBiG filtering: justify the 45% operational cutoff and clearly state that it is not an antiSMASH-recommended threshold.

• Several claims in the response to Reviewer 3 are not reflected consistently in the manuscript:

  • The response says antiSMASH 8.0, while the manuscript says 8.0.4.

  • The response refers to Table S4 for revised GCF classification, while the manuscript’s GCF distribution table appears to be Table S5.

  • The response refers to Figure 6 for uncharacterized GCFs, but Figure 6 is a synteny figure; the relevant figure appears to be Figure 8.

  • The response states that terminology and Discussion sections were substantially revised, but several inconsistencies and nomenclature errors remain.

  •  

  •  

 

• Abstract: “quasietatin S1” should likely be “squalestatin S1.”

• Section 2.2 and elsewhere: “Diapothe” should be “Diaporthe.”

• Discussion: “pathwaysimilarity” should be “pathway similarity.”

• “eninatin” / “enantian” should be “enniatin.”

• “chrysogene” should likely be “chrysogine.”

• “chaetol-ivacine” should likely be “chaetolivacine.”

• “armilly orsellinate” should likely be “armillyl orsellinate.”

• “agnstin A/B” should likely be “agnestin A/B.”

• Table 1: correct “D. cf. heveae 1,” “Ge-nome,” “Scaf-folds,” and “GC con-tent.”

• References: check formatting errors such as “Pesoonia” (should be Persoonia), “DAWSON” (all caps), and “Squalestatings.”

• Supplementary Material URLs are placeholders; ensure final links and table descriptions are complete.

• In Section 3.3, hybrid/multiclass BGCs are said to be shown in Figure 3, but Figure 3 is a Pfam domain heatmap. This likely should refer to Table 2 and/or Figure 5.

• Figure 1 appears to lack a proper figure label in the extracted text.

• Figure 7 uses asterisks to denote key biosynthetic genes, but the legend does not define which genes are marked or why.

• Figure 4 lacks a clear color legend identifying which bar corresponds to which isolate.

Author Response

Please see the attachment

Author Response File: Author Response.pdf

Reviewer 4 Report

The authors have thoroughly addressed all the previous comments and suggestions. As a result, the manuscript has become significantly more coherent, structured, and scientifically robust. The comparative genomic analysis and biosynthetic gene cluster family profiling of these Diaporthe endophytes are well-executed and provide valuable insights into their specialized metabolism.

I consider that the manuscript is now highly suitable for publication in the Journal of Fungi, subject to one minor technical correction described below.

Table 1 and Section 2.1: In Table 1, the authors provide NCBI/JGI IDs for the four analyzed genomes. While the genome data is publicly available, only one strain (Diaporthe sp. HANT25) is deposited in NCBI GenBank, whereas the other three are hosted on the JGI portal (MycoCosm). To improve data accessibility for readers and avoid confusion during data retrieval, I suggest adding a brief note in the text, a footnote to Table 1, or an update to the Data Availability Statement explicitly stating that these three assemblies (Diainc1, DiaLGMF1633, and Diavoc1) must be retrieved from the JGI Data Portal / MycoCosm repository rather than GenBank.

Author Response

Please see the attached file

Author Response File: Author Response.pdf