Functional Annotation Workflow for Fungal Transcriptomes

Round 1

Reviewer 1 Report

This manuscript presents a timely and methodologically robust fungal-specific functional annotation workflow that addresses a key gap in transcriptome analysis for non-model fungi and transcriptome-only datasets. The study is well written and convincingly demonstrates improved annotation coverage and functional resolution using both RNA-seq and Iso-Seq data. Overall, this manuscript makes a valuable contribution to fungal genomics, and I suggest accepting it with minor revisions to enhance clarity and reproducibility.

Minor Comments and Suggestions

• A short discussion on how frequently fungal-specific databases (e.g., FungiDB) are updated and how users can customize or update databases within the workflow, would strengthen practical utility.
• Clarify the rationale for prioritizing human, mouse, and budding yeast proteins in homology searches alongside fungal databases.

• Please briefly justify:
• The relatively permissive ggsearch36 e-value cutoff (E = 0.1).
• The unusually stringent differential expression threshold (padj < 1 × 10⁻¹¹) used for Lentinula edodes.
• Even a short rationale (e.g., dataset size, false-positive control) would be sufficient.

• Since transcripts are treated as independent units, it would be helpful to explicitly acknowledge potential implications related to:
• Isoform redundancy
• Functional enrichment bias
• A short clarification would improve transparency.

• The low rate of developmental stage-specific annotation (0.13%) is reported transparently. Please clarify whether this reflects biological reality, stringent thresholds, or current database limitations.
• The comparison between 14,078 reference transcripts and 227,580 assembled transcripts in L. edodes is striking. Please briefly clarify how many transcripts remained after expression and coding-region filtering to avoid the impression of excessive inflation.
• The limitation related to Iso-Seq read depth is appropriately noted; a brief mention of how hybrid short-read and long-read approaches could mitigate this would strengthen the discussion.
• The discussion of host-associated GO terms in Phakopsora pachyrhizi is insightful. Clarifying whether additional host-filtering steps were tested or considered would improve clarity.

This manuscript presents a timely and methodologically robust fungal-specific functional annotation workflow that addresses a key gap in transcriptome analysis for non-model fungi and transcriptome-only datasets. The study is well written and convincingly demonstrates improved annotation coverage and functional resolution using both RNA-seq and Iso-Seq data. Overall, this manuscript makes a valuable contribution to fungal genomics, and I suggest accepting it with minor revisions to enhance clarity and reproducibility.

Minor Comments and Suggestions

• A short discussion on how frequently fungal-specific databases (e.g., FungiDB) are updated and how users can customize or update databases within the workflow, would strengthen practical utility.
• Clarify the rationale for prioritizing human, mouse, and budding yeast proteins in homology searches alongside fungal databases.

• Please briefly justify:
• The relatively permissive ggsearch36 e-value cutoff (E = 0.1).
• The unusually stringent differential expression threshold (padj < 1 × 10⁻¹¹) used for Lentinula edodes.
• Even a short rationale (e.g., dataset size, false-positive control) would be sufficient.

• Since transcripts are treated as independent units, it would be helpful to explicitly acknowledge potential implications related to:
• Isoform redundancy
• Functional enrichment bias
• A short clarification would improve transparency.

• The low rate of developmental stage-specific annotation (0.13%) is reported transparently. Please clarify whether this reflects biological reality, stringent thresholds, or current database limitations.
• The comparison between 14,078 reference transcripts and 227,580 assembled transcripts in L. edodes is striking. Please briefly clarify how many transcripts remained after expression and coding-region filtering to avoid the impression of excessive inflation.
• The limitation related to Iso-Seq read depth is appropriately noted; a brief mention of how hybrid short-read and long-read approaches could mitigate this would strengthen the discussion.
• The discussion of host-associated GO terms in Phakopsora pachyrhizi is insightful. Clarifying whether additional host-filtering steps were tested or considered would improve clarity.

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Reviewer 2 Report

The inclusion of new bioinformatics tools to the vast list of already available programs is always welcome. However, for the sake of clarity and to avoid redundancy, it is relevant to clearly state their purpose and limitations. Having said that, the scientific community is characterized by adapting available tools to solve atypical problems. This is the case of annotation tools for fungal transcriptomes. In a strict sense, there is no such dedicated platform for fungi. Instead, tools developed for the analysis of other organisms are used. The manuscript clearly mentions this gap to be filled. However, it fails in describing why non-fungus-specific tools fail to fill this gap or the bias generated during data analysis. This is the foundation to justify the development of this kind of bioinformatics tool. Otherwise, it seems a redundant algorithm that provides not much to the specialized community. 

On the other hand, the manuscript should compare the performance of this new proposal with already established algorithms, regardless of whether they are fungus-specific or not. This comparison should provide data to support the higher analytical power of the new tool. Finally, it is suggested that the analysis will be performed with transcriptomic data of organisms of different taxonomic groups, to provide convincing data about its robustness.

The inclusion of new bioinformatics tools to the vast list of already available programs is always welcome. However, for the sake of clarity and to avoid redundancy, it is relevant to clearly state their purpose and limitations. Having said that, the scientific community is characterized by adapting available tools to solve atypical problems. This is the case of annotation tools for fungal transcriptomes. In a strict sense, there is no such dedicated platform for fungi. Instead, tools developed for the analysis of other organisms are used. The manuscript clearly mentions this gap to be filled. However, it fails in describing why non-fungus-specific tools fail to fill this gap or the bias generated during data analysis. This is the foundation to justify the development of this kind of bioinformatics tool. Otherwise, it seems a redundant algorithm that provides not much to the specialized community. 

On the other hand, the manuscript should compare the performance of this new proposal with already established algorithms, regardless of whether they are fungus-specific or not. This comparison should provide data to support the higher analytical power of the new tool. Finally, it is suggested that the analysis will be performed with transcriptomic data of organisms of different taxonomic groups, to provide convincing data about its robustness.

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Round 2

Reviewer 2 Report

The authors properly addressed my comments.

The authors properly addressed my comments.