Review Reports - Tree Fruit and Nut Crops at the Dawn of the Pangenomic Era

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

The review offers a timely synthesis on long-read sequencing, pangenomes, and integrative ‘omics’ for tree fruit and nut crops. The narrative is generally clear and well-organized, with strong didactic value. However, (i) Several claims would benefit from tighter sourcing and quantitative precision; (ii) terminology/formatting inconsistencies (species names, units, hyphenation artifacts) recur; (iii) Table 1 is useful but uneven in detail and formatting; (iv) the paper would benefit from a short “literature search strategy/selection” paragraph (even for a narrative review) to strengthen transparency.

Abstract: [L8–L27] The abstract effectively motivates the topic but uses broad statements (“billions of dollars,” “billions of humans”) without a quantitative anchor or year; consider adding a specific global value and year from FAO or similar (keep consistent with Sections 1 and refs). Also briefly define “pangenome graphs” in one clause the first time they are mentioned.
Suggestion: “Tree fruit and nut crops contribute ~$X billion (year) to the global economy and provide Y% of fruit intake…”; “Pangenome graphs—graph-based references capturing core and dispensable genomic variation—…”

Keywords: [L28–L29] Consider adding “long-read sequencing,” “Hi-C,” “graph pangenome,” “GWAS,” “structural variants,” and “genomic selection” to improve retrieval. Group general vs. technical terms.

Introduction: [L31–L39] “Olea europea” → Olea europaea (spelling). Check others similarly. The “74% of human fruit intake” claim needs precision (population scope, year, data source) to avoid over-generalization. Provide a clear numeric citation inline.

Question: What population/timeframe underlies the 74% estimate? Is it global or regional?

Table 1: [pp.3–5 around L80–L81 onward] The table is valuable but uneven: trait names vs. goals vs. examples vary in specificity; units and typography are inconsistent (e.g., “100mM NaCl” → 100 mM NaCl; add non-breaking space). Consider adding columns for “Species/cultivar,” “Evidence type (QTL/GWAS/editing/field trial),” “Effect size/phenotypic impact,” and “Readiness level (research/breeding/market).” If space is tight, convert to Supplementary Table with full details and keep a streamlined in-text table. (See also examples cited around [L234–L251] and the following section header.)

Second- vs third-generation sequencing: [L11–L36 on p.10] Great historical perspective. Please standardize accuracy figures and flow-cell/chemistry names (e.g., ONT R9/R10.x; PacBio CLR vs HiFi/CCS), and ensure that the accuracy ranges and cost statements reflect the present year (the text is already cautious but could be tightened). If newer chemistries are out of scope, add a sentence acknowledging rapid evolution and anchoring your numbers to a date.
Suggestion: Add a one-line “as of 2024/2025” qualifier where you report accuracy.

Graph pangenomes: [L13–L31 & L41–L47 on p.10] You nicely introduce PacBio/ONT and pangenomes; consider adding a one-sentence pros/cons of Minigraph vs. Minigraph-Cactus vs. PGGB (e.g., reference bias vs. scalability vs. lossless graphs) and a brief note on compute/memory considerations for tree genomes. This helps practitioners choose an approach.

Reference-guided pitfalls and “reference bias”: The narrative highlights reference bias well in earlier sections; consider one concrete tree-crop example (e.g., missed SVs in a widely used reference) and a graph-based re-analysis benefit to make it tangible for breeders. (You allude to this later; a short example here would improve flow.)

Transcriptomics & pantranscriptomes: [L388–447] The section is strong; to aid end-users, add two or three concrete decision scenarios (e.g., stress-responsive isoforms guiding rootstock choice; fruit quality pathways refining selection indices). For pantranscriptomes, only Pyrus pyrifolia is listed—flag whether other efforts are emerging and why a pantranscriptome may be the cost-effective first step before a full pangenome. (context on conclusions/flow)

Phenomics: [L451–511, esp. L486–506] Excellent coverage of UAV/LIDAR/controlled environments. Consider a schematic figure (pipeline from plot to features/traits to models to breeding decisions) and a boxed checklist (sensor, scale, trait, throughput, cost, validation). Also ensure unit/style consistency (e.g., Fv/Fm italicization, spaces in units, capitalization of trademarked tools).

In vitro phenomics: [L500–511] You correctly caution about ecological validity; consider adding one more example where in vitro diverged from field response to emphasize limits of extrapolation.

Conclusion: [L512–L523] The conclusion can be strengthened by three concrete, near-term priorities, e.g.:
(i) Genus-level graph pangenomes for two under-served nut crops;
(ii) Standardized phenomics datasets/ontologies for top five traits (color/texture/drought/cold/disease);
(iii) Breeder-ready toolchains integrating graph-aware variant calling + genomic selection.
Tie these to expected impacts (e.g., time-to-variety reduction or marker portability).

Administrative & formatting points.

Title block & affiliations: [L2–L7] Remove “Review 1” (journal placeholder); check email hyphenation line-break (“junelab-…”) and spacing around semicolons.
Nomenclature: [L36–L39] Ensure all scientific names are italicized and correctly spelled (e.g., Olea europaea, Coffea arabica, Theobroma cacao).
Broken words/hyphenation artifacts in tables: Family names split (e.g., “Juglandaceae,” “Anacardiaceae”) in Table 2/page 9–10; correct encoding/line-wrap to Juglandaceae and Anacardiaceae.
Units & spacing: Standardize to SI spacing (100 mM, 20 years, >99% consistently).
Data availability: [L526–L531] Statement is fine for a narrative review; consider adding a link to any code/notebooks used for literature collation (if any) to enhance transparency.
[L3] Tree Fruit and Nut Crops at the Dawn of the Pangenomic Era → consider adding a scope cue: “…Era: Long-Read Sequencing, Graph Pangenomes, and Phenomics” to improve discoverability.
[L15–L17] “Limitations exist, however…” → tighten to: “However, reference bias and poor marker portability limit applicability.”
[L21–L23] Define “genus-scale” briefly or add parenthetical example (e.g., Malus, Citrus).
[L33–L39] Replace commas around examples with italicized binomials and ensure “Olea europaea.” Consider adding use classes consistently (fresh, oilseed, beverage, confectionery).
[Table 1; multiple lines around L80–L120] Normalize verb tenses in “Trait goal” (present infinitive) and ensure evidence type is explicit (mutant, transgenic, QTL, GWAS, MAS).
[L11–L31 on p.10] Where you report accuracy/cost, anchor with “as of 2024/2025,” and move platform descriptions (CLR vs HiFi; ONT pore chemistry) into a concise Box 1 for clarity.
[L27–L31 on p.15] The Conclusion opening repeats earlier phrasing; consider compressing by ~15–20% and redirecting to actionable recommendations (see Major comment 11).
[References section start, p.16] Ensure uniform style (journal titles, DOIs, year formatting) and check for duplicate author names (line shows “Gschwantner … Schadauer … Gschwantner … Schadauer”; confirm correctness).

Questions

Scope & novelty. Which two to three advances since 2023 do you view as most transformative for tree pangenomics (tools, reference sets, or breeding applications)? Can you highlight concrete breeding outcomes (e.g., trait improvement enabled by SV detection or graph-aware GWAS)?
Graph pangenome adoption. For breeders with limited compute, what minimum viable pipeline (graph tool, coverage, sample number) would you recommend for a first genus-level effort?
Marker portability. Can you include an example where a graph-aware variant call rescued a previously non-transferable marker in a tree crop context?
Phenomics integration. Which phenotyping modalities (UAV multispectral, LIDAR, proximal sensors) best align with genomics for top traits (color, texture, disease, abiotic stress)? A short trait × sensor matrix would be valuable.
Pantranscriptomes vs pangenomes. For resource-limited programs, when is a pantranscriptome the optimal first step? Can you add guidance on sample sizes and tissue/stage selection?
Regulatory context. You mention gene editing policies; could you add a brief table or paragraph summarizing current regulatory stances (US, EU, Japan) specific to fruit/nut crops and implications for field deployment?
Data standards. Would you advocate specific metadata/ontology standards (MIxS-Plant, Crop Ontology) to ensure FAIR phenomics + genomics integration for perennial crops?

Suggestion

Box 1 (Glossary): graph pangenome, core vs dispensable genome, HiFi/CCS vs CLR, SV classes (DEL/INS/INV/DUP/BND), GWAS vs GS.
Figure 1: Decision tree: from research question → sequencing strategy → assembly/pangenome → variant calling (graph-aware) → trait modeling → breeding.
Supplementary Table: Curated list of public tree crop references/pangenomes with assembly metrics and links.

Author Response

Reviewer 1

Abstract: [L8–L27] The abstract effectively motivates the topic but uses broad statements (“billions of dollars,” “billions of humans”) without a quantitative anchor or year; consider adding a specific global value and year from FAO or similar (keep consistent with Sections 1 and refs). Also briefly define “pangenome graphs” in one clause the first time they are mentioned.
Suggestion: “Tree fruit and nut crops contribute ~$X billion (year) to the global economy and provide Y% of fruit intake…”; “Pangenome graphs—graph-based references capturing core and dispensable genomic variation—…”

Response: The opening sentence of the abstract has been modified for specificity based on FAOSTAT data, see changes at lines 8-10. As FAOSTAT data contains non-specific categories (e.g., Other nuts (excluding wild edible nuts and groundnuts), in shell, n.e.c.), limiting claims to an inexact “at least” estimate based on available data is preferred for clarity.

The manuscript has been changed so that “pangenome graph” has been defined in its first instance, see line 21-23.

Keywords: [L28–L29] Consider adding “long-read sequencing,” “Hi-C,” “graph pangenome,” “GWAS,” “structural variants,” and “genomic selection” to improve retrieval. Group general vs. technical terms.

Response: Two keywords have been added “genomic selection” and “GWAS”, see changes at line 32. Horticulturae imposes a limit of 10 keywords. The ordering of keywords has also been changed.

Introduction: [L31–L39] “Olea europea” → Olea europaea (spelling). Check others similarly. The “74% of human fruit intake” claim needs precision (population scope, year, data source) to avoid over-generalization. Provide a clear numeric citation inline.

Question: What population/timeframe underlies the 74% estimate? Is it global or regional?

Response: The cited paper (McMullin et al., 2019) supplies the estimate as a global figure for 2016, based on analysis of FAOSTAT data from 2016. This analysis included “tree-like” monocot crops, and has been accordingly qualified. A year qualifier has been added, as well. See line 39.

Table 1: [pp.3–5 around L80–L81 onward] The table is valuable but uneven: trait names vs. goals vs. examples vary in specificity; units and typography are inconsistent (e.g., “100mM NaCl” → 100 mM NaCl; add non-breaking space). Consider adding columns for “Species/cultivar,” “Evidence type (QTL/GWAS/editing/field trial),” “Effect size/phenotypic impact,” and “Readiness level (research/breeding/market).” If space is tight, convert to Supplementary Table with full details and keep a streamlined in-text table. (See also examples cited around [L234–L251] and the following section header.)

Response: Typographic errors within this table have been corrected, see changes to Table 1. As taxon is already listed all examples, a separate column would be redundant. Some degree of unevenness cannot be remedied when discussing traits and phenotypes which are very different in nature (e.g., temporal traits like precocity vs. physical traits like fruit color). The suggested evidence type column, becomes unwieldy in light of this fact – comparing evidence types between studies for marker development, field trials resulting from traditional breeding crosses, gene editing experiments for reverse genetics understanding, etc. becomes more confusing than illuminating. As noted in the caption, the list is non-exhaustive, and it is intended to be illustrative in nature.

Second- vs third-generation sequencing: [L11–L36 on p.10] Great historical perspective. Please standardize accuracy figures and flow-cell/chemistry names (e.g., ONT R9/R10.x; PacBio CLR vs HiFi/CCS), and ensure that the accuracy ranges and cost statements reflect the present year (the text is already cautious but could be tightened). If newer chemistries are out of scope, add a sentence acknowledging rapid evolution and anchoring your numbers to a date.
Suggestion: Add a one-line “as of 2024/2025” qualifier where you report accuracy.

Response: A clarifying statement “as of 2025” has been added at lines 299, in order to anchor accuracy estimates to a year. Nanopore flow cell references have been standardized (Nanopore R… flow cells), see changes at lines 291 and 292.

Graph pangenomes: [L13–L31 & L41–L47 on p.10] You nicely introduce PacBio/ONT and pangenomes; consider adding a one-sentence pros/cons of Minigraph vs. Minigraph-Cactus vs. PGGB (e.g., reference bias vs. scalability vs. lossless graphs) and a brief note on compute/memory considerations for tree genomes. This helps practitioners choose an approach.

Response: Some additional discussion of the tradeoffs between these most popular pangenome construction tools has been added, see changes from lines 383-395. The authors see these tools as a continuum of trade-off between performance and accuracy, with Minigraph being the most performance focused and PGGB producing the most information-rich graphs.

Reference-guided pitfalls and “reference bias”: The narrative highlights reference bias well in earlier sections; consider one concrete tree-crop example (e.g., missed SVs in a widely used reference) and a graph-based re-analysis benefit to make it tangible for breeders. (You allude to this later; a short example here would improve flow.)

Response: Discussion of pangenome graph-based findings before full introduction of pangenomes further on in the paper may unnecessarily confuse the reader, interrupting the logical flow of the review. At the end of section 2, a brief recognition of graph-based genomic representations has been added, see changes at lines 221-222.

Transcriptomics & pantranscriptomes: [L388–447] The section is strong; to aid end-users, add two or three concrete decision scenarios (e.g., stress-responsive isoforms guiding rootstock choice; fruit quality pathways refining selection indices). For pantranscriptomes, only Pyrus pyrifolia is listed—flag whether other efforts are emerging and why a pantranscriptome may be the cost-effective first step before a full pangenome. (context on conclusions/flow)

Response: As identified in the review, transcriptomics is typically a method for discovering trait-gene relationships and hypothesis development. Explicit examples in literature of breeding programs themselves using transcriptomic data as selection criteria are therefore lacking, and so this cannot be added. The main identified benefit of the pantranscriptome over the pangenome, resource efficiency due to the small size and relatively simple assembly of transcriptomes, has already been identified in the manuscript. As such, the authors see no need to belabor the point.

Phenomics: [L451–511, esp. L486–506] Excellent coverage of UAV/LIDAR/controlled environments. Consider a schematic figure (pipeline from plot to features/traits to models to breeding decisions) and a boxed checklist (sensor, scale, trait, throughput, cost, validation). Also ensure unit/style consistency (e.g., Fv/Fm italicization, spaces in units, capitalization of trademarked tools).

Response: A figure representation connecting phenomics into the pipeline for trait improvement has been added, see new graphical abstract. While great strides have been made in phenomics, most recent improvements have been improvements upon relatively straightforward sensors (e.g., multispectral cameras, but now integrated into UAVs, automated controlled environment systems, etc.), so narrative discussion of these advancements is more suited to this than a matrix or checklist.

Units have been checked for appropriate typography.

In vitro phenomics: [L500–511] You correctly caution about ecological validity; consider adding one more example where in vitro diverged from field response to emphasize limits of extrapolation.

Response: An additional example of divergence between in vitro and field trial responses in quince has been added. See lines 556-560.

Conclusion: [L512–L523] The conclusion can be strengthened by three concrete, near-term priorities, e.g.:
(i) Genus-level graph pangenomes for two under-served nut crops;
(ii) Standardized phenomics datasets/ontologies for top five traits (color/texture/drought/cold/disease);
(iii) Breeder-ready toolchains integrating graph-aware variant calling + genomic selection.
Tie these to expected impacts (e.g., time-to-variety reduction or marker portability).

Response: The authors have considered these points and have modified the conclusion section, see changes at lines 580-599. The authors believe that the currently sparse and poorly standardized landscape of pangenome analysis tools and formats is the largest barrier to this new form of genomics analysis. While it is inevitable that genus-scale pangenomes will proliferate with the recent progress is sequencing and assembly technology (just as reference genomes proliferated in the 2010s), a potentially continuing lag in the tools to utilize these resources would be damaging. Additionally, opportunities for applying pangenomic analysis to orphan crops and region-specific crops in the context of decreasing costs has been addressed. The need for standardization of phenomic metadata is also now addressed in conclusion.

Administrative & formatting points.

Title block & affiliations:[L2–L7] Remove “Review 1” (journal placeholder); check email hyphenation line-break (“junelab-…”) and spacing around semicolons.

Response: The article type is not a placeholder and is a necessary aspect of the journal format. A quick review of Horticulturae articles will show the article type (in this case, a review) one line ahead of the title. A tab has been inserted to preserve the email formatting. Extraneous spaces have been removed. See changes at lines: 4-5.

Nomenclature:[L36–L39] Ensure all scientific names are italicized and correctly spelled (e.g., Olea europaea, Coffea arabica, Theobroma cacao).

Response: Italicization and spelling have been checked. Olea europaea appears to be the primary misspelling, and one paragraph has non-italicized species names. This has been rectified.

Broken words/hyphenation artifacts in tables: Family names split (e.g., “Juglandaceae,” “Anacardiaceae”) in Table 2/page 9–10; correct encoding/line-wrap to Juglandaceae and Anacardiaceae.

Response: The table has been modified to avoid line breaks within family names. See updated Table 2.

Units & spacing: Standardize to SI spacing (100 mM, 20 years, >99% consistently).

Response: The 100 mM spacing has been altered, see changes to Table 1.

Data availability:[L526–L531] Statement is fine for a narrative review; consider adding a link to any code/notebooks used for literature collation (if any) to enhance transparency.

Response: No code was used in order to review the literature for this review.

[L3]Tree Fruit and Nut Crops at the Dawn of the Pangenomic Era → consider adding a scope cue: “…Era: Long-Read Sequencing, Graph Pangenomes, and Phenomics” to improve discoverability.

Response: The use of keywords effectively improves discoverability, and the comment about adding certain keywords has been implemented to the benefit of the review. Doubling the length of the title to include these keywords again likely will not contribute much to improving discoverability, while reducing the focus of the title.

[L15–L17] “Limitations exist, however…” → tighten to: “However, reference bias and poor marker portability limit applicability.”

Response: This line sentences have been edited as requested for clarity, see changes at line 16.

[L21–L23] Define “genus-scale” briefly or add parenthetical example (e.g., Malus, Citrus).

Response: Parenthetical examples have been added, see line 24.

[L33–L39] Replace commas around examples with italicized binomials and ensure “Olea europaea.” Consider adding use classes consistently (fresh, oilseed, beverage, confectionery).

Response: Spelling of Olea europaea has been fixed, see line 42, and elsewhere in the review. The proposed use classes provided by the reviewer do not appear any more consistent than the ones currently in the review, and in fact may be more misleading to readers, particularly as the primary tree fruit oil crop, olives, are used for their oil rich fruits, not seeds.

[Table 1; multiple lines around L80–L120] Normalize verb tenses in “Trait goal” (present infinitive) and ensure evidence type is explicit (mutant, transgenic, QTL, GWAS, MAS).

Response: All verbs in the trait goal column are in the same tense. As discussed above, the heterogeneity of examples makes adding an evidence type column undesirable.

[L11–L31 on p.10] Where you report accuracy/cost, anchor with “as of 2024/2025,” and move platform descriptions (CLR vs HiFi; ONT pore chemistry) into a concise Box 1 for clarity.

Response: A year anchor has been added. See changes to line 299.

[L27–L31 on p.15] The Conclusion opening repeats earlier phrasing; consider compressing by ~15–20% and redirecting to actionable recommendations (see Major comment 11).

Response: The conclusion has been significantly revised in response to other comments, and includes recommendations, particularly regarding the need for pangenome analysis tools. See changes to lines 580-599.

[References section start, p.16] Ensure uniform style (journal titles, DOIs, year formatting) and check for duplicate author names (line shows “Gschwantner … Schadauer … Gschwantner … Schadauer”; confirm correctness).

Response: References have been screened for errors and revised. See the updated section at line 612 onwards.

Questions

Scope & novelty. Which two to three advances since 2023 do you view as most transformative for tree pangenomics (tools, reference sets, or breeding applications)? Can you highlight concrete breeding outcomes (e.g., trait improvement enabled by SV detection or graph-aware GWAS)?

Response: The improvement of error correction methods enabling high-quality genome assembly using only Nanopore data has been particularly transformative, and greater attention has been called to this, see changes to lines 321-328. Gapless, telomere-to-telomere assemblies using a single platform and method (ONT simplex) greatly increases the accessibility of genomics and pangenomics, particularly critical for less cultivated crops. For tree fruit crops, which often have lower economic impact than other crop classes, like cereal crops, this is particularly noteworthy. The maturation of PGGB and Minigraph-Cactus, which have now become relatively user-friendly tools for pangenome graph generation, has been another major transformation; while the first version on github for these predates 2023, these were first academically published in 2023 and 2024, respectively, and have been improved greatly since their first public version. This is a step towards an easy to replicate, end-to-end pangenome analysis pipeline, which can empower breeders and a wider variety of research groups. Putting these technical improvements into context is a motivating factor for this review. As this review primarily concerns emerging technologies and analysis, and tree crop pangenomes are very new, there are no concrete breeding outcomes as of yet. The juvenile stage of development for most tree fruit and nut crops is longer than the span of time between the publication of the first tree graph pangenomes and today.

Graph pangenome adoption. For breeders with limited compute, what minimum viable pipeline (graph tool, coverage, sample number) would you recommend for a first genus-level effort?

Response: A genus-level pangenome may be inadvisable for those with limited computational resources. Pre-existing knowledge about the population genetics of the taxon in question would likely be necessary for an informed decision regarding sample selection. A genus with a cosmopolitan distribution will obviously have different requirements from a genus with a relatively narrow geographic scope, or a genus which has only recently diverged. As such, specific recommendations may be better suited to in-depth discussion in a methods paper, rather than in this review.

Marker portability. Can you include an example where a graph-aware variant call rescued a previously non-transferable marker in a tree crop context?

Response: To the authors’ knowledge, a previous marker has not been “rescued” as the result of pangenomic analysis, so much as new genetic variants linked to a specific phenotype have been discovered. For example, the Malus pangenome created a new marker related to apple scab resistance, which is intended to be completely new, rather than confirmation or refutation of a previous marker. The development of a new marker in Malus has been added explicitly, see lines 377-378.

Phenomics integration. Which phenotyping modalities (UAV multispectral, LIDAR, proximal sensors) best align with genomics for top traits (color, texture, disease, abiotic stress)? A short trait × sensor matrix would be valuable.

Response: As discussed within the phenomics section, UAV based systems are of particular interest in tree fruit and nut crops, due to the large spatial scales involved in tree cultivation, and a more explicit statement of this has been added at Line 536-538. As discussed above, the authors find a table or matrix to be unsuited to appropriately describing the recent advances in phenomics relevant to tree fruit and nut crops.

Pantranscriptomes vs pangenomes. For resource-limited programs, when is a pantranscriptome the optimal first step? Can you add guidance on sample sizes and tissue/stage selection?

Response: Such recommendations will be highly specific to the trait of interest and crop. Within the review, the relative advantages and disadvantages of pantranscriptome are discussed. Further broad recommendations would be trivial (e.g., “sequence RNA from the fruit if the trait is fruit specific”), while specific recommendations would require a purpose tailored methods paper (e.g., “assuming intraspecific genetic divergence of X%, Y samples will be needed to capture Z% of variants”).

Regulatory context. You mention gene editing policies; could you add a brief table or paragraph summarizing current regulatory stances (US, EU, Japan) specific to fruit/nut crops and implications for field deployment?

Response: It is broadly understood that the process is regulated in the EU and Japan, and the end products is regulated in the US and the countries that follow the US in their regulatory standards. Since the regulations are beyond the scope of the review, that has not been discussed here.

Data standards. Would you advocate specific metadata/ontology standards (MIxS-Plant, Crop Ontology) to ensure FAIR phenomics + genomics integration for perennial crops?

Response: Discussion regarding metadata has been added to the review to underline the importance of FAIR omics data, with MIAPPE as a leading standard. See changes at lines 565-578. Given the ability for proper metadata annotation to greatly increase the efficiency of functional genomics research, this is a particularly valuable addition. As there a number of competing ontologies for traits, environmental conditions, experimental designs, etc. specific recommendations cannot be given for all (which would likely constitute its own review). Given ongoing issues with non-standard metadata annotation, improving usage of metadata standards is key, and the specific ontology used is secondary to this goal. In addition, discussion of pangenome formats and the need for standardization has been added at lines 397-406.

Suggestion

Box 1 (Glossary):graph pangenome, core vs dispensable genome, HiFi/CCS vs CLR, SV classes (DEL/INS/INV/DUP/BND), GWAS vs GS.

Response: As each of these terms have been defined in the text, duplicated definitions would likely be unhelpful to the reader. As SV classes are not discussed at length in this review, definition of specific SV abbreviations will not be useful in our opinion.

Figure 1: Decision tree: from research question → sequencing strategy → assembly/pangenome → variant calling (graph-aware) → trait modeling → breeding.

Response: In order to graphically represent the connections between concepts within this review, we have incorporated a graphical abstract, which fits genomics, transcriptomics, and phenomics into the process of crop genetic improvement.

Supplementary Table: Curated list of public tree crop references/pangenomes with assembly metrics and links.

Response: At the genus-scale, only two pangenomes have been published for the crop types discussed, and their basic statistics have been discussed. At the species level, the definition of “pangenome” can become problematic. Publications including only a small number of genomic assemblies may be titled “pangenomes”, while representing a biased or incomplete sampling of a taxon of interest. As such, the authors deliberately did not include such a table in the review, as the result would likely be excessively subjective and not sufficiently illuminating.

Author Response File: Author Response.pdf

Reviewer 2 Report

Comments and Suggestions for Authors

In my opinion, the title does not fully reflect the content of the manuscript; in fact, the pangenomic analysis is only addressed in a relatively small part of the article. I suggest that instead of pangenomics, something like "Contribution of massive sequencing techniques to breeding of tree fruit and nut crops" be mentioned.
The objectives of the study are missing at the final part of the Introduction section. The objectives are necessary, as they provide a useful guide for drawing conclusions.
I suggest the creation of a new subsection from line 332, with the name: 4. Pangenomic analysis, which implies renumbering of subsequent subsections.
It would be very illustrative to include examples of pangenomic graphs as those mentioned in lines 357-358 in order to enable the reader for a better understanding of the subject.
It is suggested to the authors to emphasize that despite recent advances, in fact, phenotyping continues to be the most challenging part in this type of studies, due to the inherent interference of the environment, which is accentuated in perennial species.
The conclusions should be rewritten, as they now look like a summary of the manuscript. Please rewrite the conclusions based on the objectives set out in the Introduction.

Author Response

Reviewer 2

Comments and Suggestions for Authors

In my opinion, the title does not fully reflect the content of the manuscript; in fact, the pangenomic analysis is only addressed in a relatively small part of the article. I suggest that instead of pangenomics, something like "Contribution of massive sequencing techniques to breeding of tree fruit and nut crops" be mentioned.

Response: While the manuscript discusses more topics than pangenomes alone, the emergence of genus-scale pangenomes for horticultural crops is the culmination of great advancements in genomic sequencing technology, and the most notable advancement in the past few years. These advancements interact with the state-of-the-art in transcriptomics and phenomics, as discussed in the paper. Since it is a matter of opinion, we believe the title is accurate.

The objectives of the study are missing at the final part of the Introduction section. The objectives are necessary, as they provide a useful guide for drawing conclusions.

Response: The end of the Introduction has been modified to make the objectives of this review explicit. See changes at lines 174-178.

I suggest the creation of a new subsection from line 332, with the name: 4. Pangenomic analysis, which implies renumbering of subsequent subsections.

Response: A new subsection titled “Pangenomic Analysis” has been added, see line 355.

It would be very illustrative to include examples of pangenomic graphs as those mentioned in lines 357-358 in order to enable the reader for a better understanding of the subject.

Response: A graphical abstract has been added to the manuscript, which includes a graphical representation of a (very short) pangenome graph. This should complement the discussion of two specific tree fruit pangenomes (for Malus and Citrus) which are explored within the review.

It is suggested to the authors to emphasize that despite recent advances, in fact, phenotyping continues to be the most challenging part in this type of studies, due to the inherent interference of the environment, which is accentuated in perennial species.

Response: Mention of environmental factors confounding the interpretation and sharing of phenotypic data has been added, see lines 570-571, in addition to the already robust discussion of the “phenotyping bottleneck” and limitations to the collection of data specific to tree crops. Additionally, discussion of metadata standards which include environmental conditions and experimental designs which may improve data reuse has been added, see lines 565-578.

The conclusions should be rewritten, as they now look like a summary of the manuscript. Please rewrite the conclusions based on the objectives set out in the Introduction.

Response: The conclusions section has been modified significantly and makes note of the objectives set forth in the introduction. See modifications at lines 580-599.

Author Response File: Author Response.pdf

Reviewer 3 Report

Comments and Suggestions for Authors

This manuscript focuses on "Tree Fruit and Nut Crops at the Dawn of the Pangenomic Era," with a topic that embodies both academic frontier and industrial practicality. It not only aligns with current cutting-edge genomic technologies such as pangenomics and third-generation sequencing but also directly addresses core industrial pain points in tree fruit and nut crop breeding, such as long breeding cycles and phenotyping bottlenecks. This work holds significant reference value for bridging basic research in this field with genetic improvement practices. The manuscript follows a clear "problem-technology-application" logical framework (introducing demands → analyzing multi-omics technologies → implementing improvement practices → concluding and prospecting) with well-structured sections. It is supported by solid data, citing numerous key literatures in the field (e.g., pangenome cases of apple and citrus) and specific improvement indicators (e.g., cultivar development duration, planting density), ensuring high credibility of arguments. However, from the perspective of "depth and guidance of a field review," the manuscript still has room for optimization in terms of technical detail connection, case expansion, and practical guidance value. Specific comments and suggestions are as follows: 1.The manuscript does not discuss genomics, transcriptomics, or phenomics in isolation but highlights the chain of "genomics + pangenomics → transcriptomic annotation → phenomic validation → improvement technology implementation" (e.g., variable genes identified by pangenomics are validated for expression patterns via transcriptomics, and then associated with traits through phenomics). This logic is consistent with the current trend of "systems biology-driven breeding." It is recommended to further emphasize this integration approach in the abstract and conclusion, clarifying the core viewpoint of "how multi-omics collaboratively address traditional breeding bottlenecks" to enhance the academic recognition of the manuscript.

2.Add quantitative data: For example, "FAO data (2024) shows that X billion people worldwide still suffer from vitamin C deficiency, while tree fruit crops contribute Y% of global vitamin C supply. However, their yield growth rate is only Z%, far below the demand from population growth." This will make the connection between nutritional gaps and crop importance more intuitive.

3. Clarify smallholder agriculture cases: For example, "Data from India’s Mahagrapes cooperative indicates that smallholders growing grapes (a tree fruit crop) have 30%-50% higher incomes than those growing rice [122]." Using specific research to support the "supporting role of tree fruit crops for smallholders" will enhance the persuasiveness of arguments.

4.Add qualifications to some terms. For example, when "third-generation sequencing" is first mentioned, supplement "(also known as long-read sequencing, with read lengths of 10 kb–1 Mb+)" to avoid misunderstandings by non-specialist readers. Distinguish between "species-level pangenomes" and "genus-level pangenomes (super-pangenomes)" for "pangenome" to prevent conceptual confusion.

5. In the transition section from "pangenomics to genetic improvement" (e.g., the beginning of Chapter 6), add "By reducing reference bias, pangenomics improves the accuracy of marker-trait associations in genomic selection, thereby providing more precise targets for CRISPR editing (e.g., the PH4 gene in citrus)." This strengthens the connection between technologies and avoids a sense of "disconnected technical accumulation."

This manuscript is a high-quality field review with a clear core framework, solid data support, and potential for publication in horticultural journals such as Horticulturae. If optimized based on the three core directions of "strengthening quantitative support, supplementing practical guidance, and optimizing logical coherence" (e.g., adding a technology selection guide, quantifying improvement effects, standardizing formats), the depth and practicality of the manuscript will be further enhanced. It will then not only serve as a "summary of field progress" but also provide a full-chain reference for researchers covering "technology selection-experiment design-breeding implementation."

Author Response

Reviewer 3

1.The manuscript does not discuss genomics, transcriptomics, or phenomics in isolation but highlights the chain of "genomics + pangenomics → transcriptomic annotation → phenomic validation → improvement technology implementation" (e.g., variable genes identified by pangenomics are validated for expression patterns via transcriptomics, and then associated with traits through phenomics). This logic is consistent with the current trend of "systems biology-driven breeding." It is recommended to further emphasize this integration approach in the abstract and conclusion, clarifying the core viewpoint of "how multi-omics collaboratively address traditional breeding bottlenecks" to enhance the academic recognition of the manuscript.

Response: The abstract and conclusion have both been modified to further underline the importance of a multi-omics approach in tree fruit and nut crop genetic improvement. See changes at lines 28-30 and 580-599.

Response: Concrete numbers have been placed on inadequate intake of Vitamins A and C globally, and the potential benefits of tree product consumption have been underlined here. See alterations at lines 54-57.

Clarify smallholder agriculture cases: For example, "Data from India’s Mahagrapes cooperative indicates that smallholders growing grapes (a tree fruit crop) have 30%-50% higher incomes than those growing rice [122]." Using specific research to support the "supporting role of tree fruit crops for smallholders" will enhance the persuasiveness of arguments.

Response: A specific instance of tree fruit cultivation was added to illuminate the profound economic effects these crops can bring to smallholder farmers, see additions at lines 65-67. Grapes are outside of the scope of this review (not fitting into the definition of “tree” outlined in the introduction), but other examples were available to contribute to this point.

Response: As readers of a plant science journal can be expected to understand the definition of a genus and a species, definition of “genus-scale pangenome” and “species-specific pangenome” would not be particularly illuminating. The definition of a pangenome and a graph pangenome are provided, as these terms are far less self-explanatory.

In the transition section from "pangenomics to genetic improvement" (e.g., the beginning of Chapter 6), add "By reducing reference bias, pangenomics improves the accuracy of marker-trait associations in genomic selection, thereby providing more precise targets for CRISPR editing (e.g., the PH4gene in citrus)." This strengthens the connection between technologies and avoids a sense of "disconnected technical accumulation."

Response: Towards the end of the discussion of pangenomes, a line more explicitly tying this chapter into the wider goal of tree fruit and nut crop genetic improvement has been added, see lines 417-421.

Author Response File: Author Response.pdf