Poly pipeline: a polyvalent spatial transcriptomics workflow validated across polyploid and diploid organisms
Curation statements for this article:-
Curated by GigaByte
Editors Assessment:
This Technical Release (software/workflow paper) presents Poly Pipeline, an open-source spatial transcriptomics workflow designed to handle the data sparsity, fragmented proprietary formats and duplicated-gene complexity that make polyploid plant genomes difficult to analyze. Carvalho et al. combine a converter that turns h5ad and GEM files into the GEF format used by Stereopy, quality filtering, Louvain, Leiden and Spatial Leiden clustering, a user-defined “anchoring” step for assigning regions, and highly variable and marker gene detection. Optional hdWGCNA co-expression networks are built on KNN-derived metacells, intended to retain low-expressed duplicated genes that would otherwise be discarded as dropouts. The pipeline was developed on a hexaploid wheat seed Stereo-seq dataset and benchmarked on rice, Arabidopsis and mouse data in different formats. Spatial Leiden gave more cohesive tissue domains than standard Leiden (15 versus 26 clusters in wheat), and the largest dataset (40,248 cells, 70,766 genes) ran in about 23.5 minutes using a peak of 31.8 GB of RAM. The benchmarking mainly shows that the workflow runs reliably and produces spatially coherent domains across organisms and formats. It does not directly test the claim that hdWGCNA better preserves duplicated-gene signatures, so that remains largely supported by argument rather than comparison with other methods. Nonetheless, with the code available on GitHub under an MIT license (with Zenodo, WorkflowHub, RRID and bio.tools entries), this provides a reproducible, HPC-ready framework that should help plant researchers working with complex genomes and multi-platform spatial datasets.
This evaluation refers to version 1 of the preprint
This article has been Reviewed by the following groups
Listed in
- Evaluated articles (GigaByte)
Abstract
Abstract Spatial transcriptomics (ST) has emerged as a transformative approach for visualizing tissue landscapes, yet it faces significant challenges regarding data standardization, sparsity, and the analysis of complex genomes, particularly polyploid plants. To address these limitations, we introduce Poly Pipeline, a robust and universal bioinformatic workflow designed to streamline analysis across diverse plant and animal genomes. The pipeline integrates a comprehensive converter for proprietary formats, clustering algorithms, and hdWGCNA co-expression networks, which indirectly preserves the expression signatures of low-expressed duplicated genes. Benchmarking across datasets from wheat, rice, Arabidopsis, and mouse demonstrated the broad applicability of the pipeline in identifying relevant clusters, showing effectiveness across diverse organisms and data types. By providing a unified and reproducible framework, Poly Pipeline addresses a critical gap in analyzing genomic redundancy, especially that related to polyploidy, and promotes FAIR data principles for the broader scientific community.
Article activity feed
-
Editors Assessment:
This Technical Release (software/workflow paper) presents Poly Pipeline, an open-source spatial transcriptomics workflow designed to handle the data sparsity, fragmented proprietary formats and duplicated-gene complexity that make polyploid plant genomes difficult to analyze. Carvalho et al. combine a converter that turns h5ad and GEM files into the GEF format used by Stereopy, quality filtering, Louvain, Leiden and Spatial Leiden clustering, a user-defined “anchoring” step for assigning regions, and highly variable and marker gene detection. Optional hdWGCNA co-expression networks are built on KNN-derived metacells, intended to retain low-expressed duplicated genes that would otherwise be discarded as dropouts. The pipeline was developed on a hexaploid wheat seed Stereo-seq dataset and benchmarked on rice, …
Editors Assessment:
This Technical Release (software/workflow paper) presents Poly Pipeline, an open-source spatial transcriptomics workflow designed to handle the data sparsity, fragmented proprietary formats and duplicated-gene complexity that make polyploid plant genomes difficult to analyze. Carvalho et al. combine a converter that turns h5ad and GEM files into the GEF format used by Stereopy, quality filtering, Louvain, Leiden and Spatial Leiden clustering, a user-defined “anchoring” step for assigning regions, and highly variable and marker gene detection. Optional hdWGCNA co-expression networks are built on KNN-derived metacells, intended to retain low-expressed duplicated genes that would otherwise be discarded as dropouts. The pipeline was developed on a hexaploid wheat seed Stereo-seq dataset and benchmarked on rice, Arabidopsis and mouse data in different formats. Spatial Leiden gave more cohesive tissue domains than standard Leiden (15 versus 26 clusters in wheat), and the largest dataset (40,248 cells, 70,766 genes) ran in about 23.5 minutes using a peak of 31.8 GB of RAM. The benchmarking mainly shows that the workflow runs reliably and produces spatially coherent domains across organisms and formats. It does not directly test the claim that hdWGCNA better preserves duplicated-gene signatures, so that remains largely supported by argument rather than comparison with other methods. Nonetheless, with the code available on GitHub under an MIT license (with Zenodo, WorkflowHub, RRID and bio.tools entries), this provides a reproducible, HPC-ready framework that should help plant researchers working with complex genomes and multi-platform spatial datasets.
This evaluation refers to version 1 of the preprint
-
AbstractSpatial transcriptomics (ST) has emerged as a transformative approach for visualizing tissue landscapes, yet it faces significant challenges regarding data standardization, sparsity, and the analysis of complex genomes, particularly polyploid plants. To address these limitations, we introduce Poly Pipeline, a robust and universal bioinformatic workflow designed to streamline analysis across diverse plant and animal genomes. The pipeline integrates a comprehensive converter for proprietary formats, clustering algorithms, and hdWGCNA co-expression networks, which indirectly preserves the expression signatures of low-expressed duplicated genes. Benchmarking across datasets from wheat, rice, Arabidopsis, and mouse demonstrated the broad applicability of the pipeline in identifying relevant clusters, showing effectiveness across …
AbstractSpatial transcriptomics (ST) has emerged as a transformative approach for visualizing tissue landscapes, yet it faces significant challenges regarding data standardization, sparsity, and the analysis of complex genomes, particularly polyploid plants. To address these limitations, we introduce Poly Pipeline, a robust and universal bioinformatic workflow designed to streamline analysis across diverse plant and animal genomes. The pipeline integrates a comprehensive converter for proprietary formats, clustering algorithms, and hdWGCNA co-expression networks, which indirectly preserves the expression signatures of low-expressed duplicated genes. Benchmarking across datasets from wheat, rice, Arabidopsis, and mouse demonstrated the broad applicability of the pipeline in identifying relevant clusters, showing effectiveness across diverse organisms and data types. By providing a unified and reproducible framework, Poly Pipeline addresses a critical gap in analyzing genomic redundancy, especially that related to polyploidy, and promotes FAIR data principles for the broader scientific community.
This paper has been published in GigaByte journal under a CC-BY license with open peer reviews: https://doi.org/10.46471/gigabyte.190
Reviewer 1. Guanjing Hu and Cong Sun
Is the code executable? Unable to test.
The authors presented Poly Pipeline as a standardized downstream analysis workflow for spatial transcriptomics (ST) data. Built around BGI's Stereopy package, it wraps the downstream analysis into a single packaged pipeline, including a proprietary-format converter (h5ad/GEM to GEF), several clustering methods (Louvain, Leiden, Spatial Leiden, and a custom "anchoring" step), highly variable and marker gene detection, and hdWGCNA-based co-expression network analysis, with benchmarking on wheat, rice, Arabidopsis, and mouse. Relative to Stereopy itself, the added value of Poly Pipeline lies in packaging the full downstream analysis and in its ease of use, rather than in algorithmic innovation. Nevertheless, because it provides a standardized and reproducible ST downstream pipeline, I recommend the paper for publication after the following revisions.
Major comment: The title, abstract, and introduction all position "overcoming polyploid complexity" as the primary innovation, but the manuscript does not engage with the core question it raises: what, specifically, does this pipeline do to address polyploid complexity that a standard diploid-oriented pipeline does not? Polyploid genomes present well-defined analytical challenges, including homeologous gene copies that share high sequence similarity, subgenome-specific expression biases, dosage effects from multiple gene copies, and the potential for read multi-mapping ambiguity. These are distinct from the data sparsity problem, which is shared across all ST platforms and organisms. The pipeline's methods rely on Stereopy's Louvain/Leiden clustering (a generic graph-based approach with no homoeolog-aware logic), hdWGCNA's metacell KNN bagging (engineered to mitigate zero-inflated expression matrices, a universal ST issue), and format conversion and standard QC (platform-compatibility benefits, not polyploid-specific). None of these components introduces allele-level or subgenome-level resolution. The pipeline does not, for instance, triage reads by subgenome, assign expression to A/B/D homeologs in wheat, or employ dosage-normalized differential expression. The Discussion's most compelling biological example, PEPC gene subgenome retention and spatial localization in wheat endosperm, is drawn from Millsteed et al. (2025)'s published biological analysis, not from a pipeline-enabled discovery. Framing it as a validation that the pipeline recapitulates known biology is fine; leaving it ambiguous implies an analytical capability the pipeline does not actually provide. Overall, the pipeline has clear value, but it does not target the key polyploid-specific methodology challenges, and the manuscript should not claim otherwise without new analysis to support it. So my recommendation for the authors is either (a) incorporate and demonstrate concrete polyploid-aware analytical modules (e.g., homoeolog-resolved expression quantification, subgenome-aware clustering, or dosage-corrected network construction), or (b) significantly downplay the polyploid-specific claim throughout the manuscript, retitle, revise the abstract, and recast the innovation as "a unified ST workflow validated on diverse organisms including polyploid plants."
Minor comments:
- Verify whether the hdWGCNA package is genuinely used in the scripts. The manuscript's wording regarding hdWGCNA should be consistent with what the code actually runs.
- The pipeline itself does not demonstrate any improvement in the quality of the raw data. I suggest revising phrases such as "mitigate sparsity and resolve duplicated gene functions" in the abstract, and "solving the sparsity" and "considerably reducing dropouts" in the discussion, to more cautious wording, or supporting them with quantitative evidence.
- All four benchmark datasets are BGI Stereo-seq, and the main workflow is built on the Stereopy package. I suggest adding at least one complete worked example on 10x Visium or another spatial transcriptomics platform, or restricting the stated "universality" to "universality within the Stereo-seq/Stereopy ecosystem."
- Further improve the language and the accuracy of the technical terminology. For example: (i) the resolution unit is described as a "nanopore radius," but Stereo-seq is based on DNA nanoball (DNB) patterned arrays, not nanopores; (ii) The statement "1 bin unit representing 10–14 microns" is inaccurate because the bin size is not specified. DNA nanoball diameter is much smaller than 10-14 microns; and (iii) "Giotto se" should be "Giotto."
Reviewer 2. Haojing Shao
This paper presents Poly Pipeline, a universal bioinformatics workflow for spatial transcriptomics optimized for polyploid organisms. It solves key issues including inconsistent proprietary data formats, severe data sparsity and duplicated gene bias in polyploid genomes. The pipeline contains multi-step modules: format conversion, quality filtering, PCA/UMAP dimensional reduction, multi-mode spatial clustering, marker gene screening and hdWGCNA co-expression network construction to reduce dropout noise. Benchmarks on wheat, rice, Arabidopsis and mouse datasets prove its efficiency and wide applicability. Spatial Leiden clustering delivers clearer tissue domain segmentation. This open-source Linux pipeline follows FAIR data principles and works for both plants and animals. Comments: Major comments:
- What is the major novelty and advantage of the Poly Pipeline software? Is the main contribution the development of a Snakemake-based framework to integrate these analyses? Most of these analyses can also be performed using existing tools without this pipeline. The authors should clearly explain the unique value and advantages of Poly Pipeline from the perspective of the analysis workflow and biological applications. 2. The amount of presented work appears insufficient. In Figure 1, the authors mention multiple analytical modules, including PCA/UMAP visualization, spatial clustering, highly variable gene (HVG) identification, hdWGCNA analysis, and network generation. However, the current results mainly focus on spatial clustering and HVG analysis. Where are the results from the other modules? I suggest adding additional results and figures, such as PCA/UMAP visualization, hdWGCNA analysis, and gene regulatory network construction, to better demonstrate the capability of the pipeline.
- The dataset used for spatial clustering in Figure 2 appears outdated, and the clustering results show obvious granularity artifacts. I recommend using a more recent high-resolution spatial transcriptomics dataset for evaluation. For example, datasets from the following resource could be considered: https://en.stomics.tech/resources/demo-data/list.html
- In Table 2, statistical analyses such as skewness, kurtosis, and the Shapiro–Wilk test should include detailed statistical values, especially the corresponding P-values. In addition, these results should be compared with existing tools. Do the results demonstrate better performance compared with other methods, or are they generally consistent with existing approaches? This comparison is important for demonstrating the advantages and reliability of Poly Pipeline.
- The title emphasizes “Poly Pipeline”; however, I did not observe any specific analyses or methodological designs targeting the unique characteristics of polyploid genomes or polyploid species. The current manuscript mainly demonstrates the application of the pipeline to polyploid species. I suggest either revising the title to better match the actual content or adding polyploid-specific analyses to highlight the unique advantages of this software.
Minor comments:
- The Introduction section does not align well with the actual analyses performed in the manuscript. I suggest revising the Introduction to focus more specifically on the problems addressed and functions implemented by Poly Pipeline.
- Is the GEF format only supported by Stereo-seq data? Can Poly Pipeline process other widely used spatial transcriptomics data formats, such as 10x Genomics Visium, Xenium datasets, h5ad files, or Seurat objects? If these formats are not currently supported, the authors should discuss this limitation in the Discussion section. Additional annotated file: https://gigabyte-review.rivervalleytechnologies.com/api/download-documents?payload=dLQpgSxI41Ksf6QFmEn6UrojoBRhxttwE1cPXu8tMOJB41W0jaAP7nLHgOCO=gp8
-
-
