TL;DR
HAPPE is a new Python-based tool that simplifies population haplotype analysis and visualization by generating editable Excel tables. It addresses the challenge of integrating detailed genotype information with broader haplotype patterns, making complex genomic data more accessible to researchers, especially those less familiar with programming.
What Problem Does This Paper Address?
As DNA sequencing costs plummet, the volume of plant genomic data is exploding. Analyzing this data, particularly for population genomics and breeding applications, requires understanding haplotypes – combinations of nearby SNPs inherited together. While many tools exist for visualizing haplotypes, they often struggle to simultaneously display the intricate details of individual SNPs alongside the overall haplotype structure. This gap makes it difficult for researchers to gain a comprehensive understanding of their data, especially when dealing with large datasets. Existing visualization tools can be image-based, hindering further analysis, or lack the ability to integrate crucial supplementary information like P-values from GWAS or sequencing depth, which are vital for detecting structural variations.
Methods
HAPPE employs an agglomerative hierarchical clustering algorithm to group samples based on their genetic similarity. The core of HAPPE's novelty lies in its output format: editable Excel tables.
Here's a breakdown of the process:
- Data Input: HAPPE accepts genotype data in VCF format, along with optional files for sample lists, colors, and sequencing depth.
- Analysis:
- Samples and variants are filtered based on user-defined criteria.
- Genotypes are converted into a numerical matrix.
- Euclidean distance is used to calculate the distance matrix between samples.
- The Ward linkage method is applied to compute the linkage matrix.
- Dynamic Tree Cut is used to define clusters and generate a phylogenetic tree.
- Normalized sequencing depth is calculated for windows (default 50 bp).
- Visualization:
- HAPPE uses the
openpyxlPython library to create Excel files.
- HAPPE uses the
- It generates a phylogenetic tree within the Excel sheet.
- Sample labels and annotations are displayed.
- Genotypes are visualized using a color-coded system within cells.
- A heatmap of sequencing depth is generated alongside the genotype matrix.
- Crucially, HAPPE allows for the parallel display of additional information, such as P-values from GWAS, gene/SNP annotations, and sequencing depth, directly within the Excel output.
Datasets
The authors used a published dataset of Setaria viridis (Mamidi et al., 2020) to test HAPPE. They downloaded reads from the NCBI (accession numbers BioProject PRJNA560514 and PRJNA265547), mapped them to the reference genome, and called SNPs. Specifically, they selected two genes, Sevir.5G085400 and Sevir.5G394700, for visualization using HAPPE.
Key Insights and Ideas
- Bridging the Gap: HAPPE effectively bridges the gap between complex genomic data and user-friendly visualization by leveraging the ubiquity and familiarity of Excel.
- Integrated Information: The ability to display multiple layers of information (phylogenetic tree, genotypes, P-values, sequencing depth, annotations) in a single, editable table is a significant advantage for comprehensive data interpretation.
- Accessibility for Non-Programmers: By outputting editable Excel files, HAPPE democratizes haplotype analysis, making it accessible to researchers without extensive programming or bioinformatics backgrounds.
- Detection of Structural Variants: The integrated sequencing depth heatmap is a powerful feature for identifying insertions/deletions or copy number variations, which can be challenging with standard SNP-based visualizations.
Key Results
HAPPE successfully generated informative and editable haplotype plots for the selected Setaria viridis genes.
- For Sevir.5G085400, HAPPE visualized haplotypes and genotypes, allowing for the detection of a deletion event in one accession (AGTWO) indicated by low sequencing depth. This deletion was further confirmed using a genome browser. The tool also enabled the visualization of geographical distribution of individuals based on their haplotypes.
- For Sevir.5G394700, HAPPE identified ten distinct haplotypes and revealed group-specific distributions, highlighting how different haplotypes were prevalent in specific subgroups (e.g., "Central_East," "West_Coast").
The tool's output is described as "editable plots consisting of cells in Excel tables," which are "user-friendly to non-programmers."
Limitations and Future Work
- Excel Limitations: The primary limitation acknowledged is the inherent constraint of Excel tables, which can only accommodate a maximum of 18,278 columns and 1,048,576 rows. This means the number of SNPs and samples that can be displayed is limited by these figures.
- Future Work: While not explicitly detailed, the paper implies that HAPPE's user-friendly nature and its ability to integrate diverse data types make it suitable for a wide range of future studies and collaborations, particularly in the context of increasing whole-genome sequencing data. The authors suggest its broad applicability in various research areas.
Summary updated