How huffordlab/nam-genomes works
Project’s Main Components
• Data Input & External Resources: Start with raw datasets (e.g., PacBio, Illumina, RNA-Seq data) and external databases (MaizeGDB, CyVerse, Gramene) as your entry points.
• Quality Control: A module consisting of scripts and pipelines for raw-data QC and assembly evaluation (e.g., quality-control folder).
• Genome Assembly & Pseudomolecule Construction: Represent the genome-assembly and agp-generation pipelines as separate components.
• Gene Prediction & Annotation: Include modules like gene-prediction (both ab-initio and evidence-based methods) and the merging of predictions.
• Population Genetics & Association Studies: Illustrate how the abc, GERP, and gwas-vcap modules fit into population genetic analysis.
• Visualization & Reporting: Highlight directories with figures, notebooks, and reporting scripts (e.g., analysis_viz in abc, publication-plots).
Relationships & Data Flow
• That processed data and assemblies feed into gene prediction/annotation and other downstream analysis modules.
• That results from individual pipelines (e.g., structural variation, methylation, pan-genomics) may have data visualization and validation steps.
• Demonstrate that many pipelines are executed via workflow managers (such as Snakefiles in the predict and simulations directories), implying an underlying dependency management layer.
Architectural Patterns & Tools
• The project is modular, with each analysis task housed in its own folder. This separation resembles a microservices-like pattern in research workflows.
• “pipeline” nature of the design: many directories contain sequential steps (e.g., “submit.sh”, “submit.json”, “run*sh” scripts) typical of batch or high-performance computing workflows.