🌱 PlantAssemblyForge

A Reproducible Nextflow DSL2 Platform for Plant Genome and Transcriptome Assembly

PlantAssemblyForge is a modular bioinformatics workflow for plant de novo genome assembly, transcriptome assembly, quality control, assembly validation, functional analysis, and interactive results visualization.

The platform combines Nextflow DSL2 workflow orchestration with established bioinformatics tools and an interactive Streamlit dashboard, providing a reproducible workflow from raw sequencing reads to interpretable assembly-quality results.

---

🖥️ Interactive Results Dashboard

PlantAssemblyForge includes an interactive Streamlit dashboard for exploring benchmark results, assembly statistics, validation metrics, k-mer distributions, transcriptome outputs, and workflow architecture.

Project Overview

The overview provides a compact summary of the benchmark dataset and major genome assembly results, including cleaned sequencing coverage, assembly size, genome fraction, and BUSCO completeness.

---

🧬 De Novo Genome Assembly

The genome workflow processes paired-end sequencing reads through quality assessment, read preprocessing, k-mer analysis, assembly, and independent assembly validation.

Paired-end FASTQ
        │
        ▼
      FastQC
        │
        ▼
       fastp
        │
        ▼
Post-trimming FastQC
        │
        ├────────────► Jellyfish k-mer analysis
        │
        ▼
      SPAdes
        │
        ├────────────► QUAST
        │
        └────────────► BUSCO

Genome Assembly Dashboard

Representative benchmark statistics include:

Metric	Result
Scaffold assembly size	107.24 Mb
Contig assembly size	107.21 Mb
Scaffold N50	8,515 bp
Contig N50	7,429 bp
Largest scaffold	56,462 bp
GC content	36.21%

---

📊 QUAST Assembly Evaluation

QUAST is used to evaluate assembly contiguity and reference-based genome recovery.

Representative results include:

Metric	Result
Genome fraction	81.892%
Duplication ratio	1.003
Scaffold NGA50	5,971 bp

QUAST outputs are retained in results_summary/genome/ to provide compact evidence of assembly performance without storing large intermediate files in Git.

---

🧬 BUSCO Completeness Assessment

BUSCO evaluates recovery of evolutionarily conserved plant genes and provides an independent measure of biological completeness.

Representative BUSCO results:

Category	Result
Complete BUSCOs	95.4%
Single-copy	93.9%
Fragmented	3.4%
Missing	1.3%

The benchmark demonstrates high recovery of conserved gene content despite fragmentation expected from a moderate-depth short-read de novo assembly.

---

🔬 k-mer Analysis

Jellyfish is used to generate the k-mer frequency spectrum before assembly.

The current benchmark uses:

k = 21

The resulting spectrum provides information about read multiplicity, sequencing error, coverage structure, and genome complexity.

The representative histogram is available at:

results_summary/genome/k21_histogram.tsv

---

🧪 Transcriptome Assembly

PlantAssemblyForge also contains modules for de novo and reference-guided transcriptome analysis.

De Novo Branch

RNA-seq
   │
   ▼
FastQC
   │
   ▼
fastp
   │
   ▼
RNA-SPAdes
   │
   ▼
Read-back mapping
   │
   ▼
rnaQUAST
   │
   ▼
TransDecoder
   │
   ▼
DIAMOND

This branch supports transcript reconstruction, assembly validation, protein prediction, and sequence-similarity-based functional analysis.

Reference-Guided Branch

RNA-seq
   │
   ▼
HISAT2
   │
   ▼
SAMtools
   │
   ▼
StringTie

The reference-guided branch provides an alternative workflow when an appropriate reference genome is available.

---

⚙️ Nextflow DSL2 Architecture

PlantAssemblyForge uses modular Nextflow DSL2 processes to separate individual bioinformatics operations from higher-level workflow logic.

Genome workflow modules include:

GENOME_FASTQC
      │
      ▼
GENOME_FASTP
      │
      ▼
GENOME_FASTQC_CLEAN
      │
      ├────────────► GENOME_JELLYFISH
      │
      ▼
GENOME_SPADES
      │
      ├────────────► GENOME_QUAST
      │
      └────────────► GENOME_BUSCO

The genome modules have also been tested using Nextflow stub execution to validate workflow connectivity without rerunning computationally expensive analyses.

---

🛠️ Technology Stack

Category	Tools
Workflow management	Nextflow DSL2
Read QC	FastQC, fastp, MultiQC
k-mer analysis	Jellyfish
Genome assembly	SPAdes
Genome validation	QUAST, BUSCO
Transcriptome assembly	RNA-SPAdes
Alignment	HISAT2, Bowtie2, minimap2
Alignment processing	SAMtools
Transcript reconstruction	StringTie
Protein prediction	TransDecoder
Functional similarity search	DIAMOND
Dashboard	Streamlit
Data analysis	Python, pandas
Visualization	Plotly, Matplotlib
Environment management	Conda

---

📁 Repository Structure

PlantAssemblyForge/
│
├── app/
│   └── app.py
│
├── conf/
│   └── genome.config
│
├── config/
│   └── samplesheet.csv
│
├── docs/
│   └── images/
│       ├── overview.png
│       ├── genome_assembly.png
│       ├── busco.png
│       └── transcriptome.png
│
├── modules/
│   ├── alignment/
│   ├── annotation/
│   ├── assembly/
│   ├── genome/
│   ├── qc/
│   └── validation/
│
├── workflows/
│   ├── genome.nf
│   └── transcriptome.nf
│
├── results_summary/
│   ├── RESULTS.md
│   └── genome/
│
├── genome_main.nf
├── main.nf
├── nextflow.config
├── environment.yml
├── CITATION.cff
├── LICENSE
└── README.md

---

🚀 Installation

Clone the repository:

git clone https://github.com/Pratik-2002-ux/PlantAssemblyForge.git
cd PlantAssemblyForge

Create the Conda environment:

conda env create -f environment.yml
conda activate plantassembly

---

▶️ Running the Genome Workflow

Validate the workflow structure using Nextflow stub execution:

nextflow -C conf/genome.config run genome_main.nf -stub-run

Run the complete genome workflow:

nextflow -C conf/genome.config run genome_main.nf

---

🖥️ Running the Dashboard

Launch the interactive results dashboard:

streamlit run app/app.py

Streamlit will provide a local browser address, typically:

http://localhost:8501

The dashboard provides dedicated views for:

• Project overview
• Genome assembly
• QUAST
• BUSCO
• k-mer analysis
• Transcriptome analysis
• Workflow architecture

---

📦 Results and Data Policy

Compact representative results are provided under:

results_summary/

Large files are deliberately excluded from Git version control, including:

• Raw FASTQ/SRA sequencing data
• Trimmed FASTQ files
• Large genome assemblies
• Reference genomes and annotations
• BUSCO databases
• Large intermediate alignment files
• Nextflow work/ directories
• Tool installations and databases

This keeps the repository lightweight while preserving the workflow implementation and representative evidence required to understand and reproduce the analysis.

---

🔁 Reproducibility

PlantAssemblyForge supports reproducibility through:

• Nextflow DSL2 workflow orchestration
• Modular process definitions
• Conda environment specification
• Explicit configuration files
• Representative benchmark outputs
• Git version control
• Versioned releases
• Citation metadata
• Stub-run workflow validation

---

📌 Current Release

PlantAssemblyForge v1.1.0

This release includes the modular workflow implementation, representative genome/transcriptome results, interactive Streamlit dashboard, and dashboard documentation.

---

📖 Citation

Citation metadata is provided through:

CITATION.cff

If PlantAssemblyForge contributes to research or analysis, please cite the repository and corresponding software release.

---

📄 License

PlantAssemblyForge is distributed under the MIT License.

See LICENSE for details.

---

👨‍💻 Author

Pratik Ramchandra Chaudhari

PlantAssemblyForge was developed as a bioinformatics workflow-development project focused on reproducible plant genome and transcriptome analysis.

---

⭐ If you find PlantAssemblyForge useful, consider starring the repository.