Probabilistic Assembly with Mycelia

Mycelia's probabilistic assembly approach combines probabilistic modelling with biological insight to deliver assemblies with confidence intervals.

🧬 What is Probabilistic Assembly?

Traditional Assembly

Traditional assemblers make deterministic decisions at each step:

Fixed k-mer size throughout assembly
Binary decisions: keep or discard sequences
Quality scores often ignored after initial filtering
Heuristic-based error correction

Probabilistic Assembly

Mycelia's probabilistic approach treats assembly as a statistical inference problem:

Dynamic k-mer progression based on data characteristics
Probabilistic path selection using quality scores
Maximum likelihood error correction
Self-optimizing parameters through machine learning

Traditional:  Reads → Graph → Heuristic Cleaning → Assembly
Probabilistic: Reads → Quality-Aware Graph → Statistical Inference → Optimal Assembly

🎯 Why Use Probabilistic Assembly?

1. High Accuracy

Preserves quality information throughout assembly
Statistically principled error correction

2. Self-Optimizing

No manual parameter tuning required
Automatically selects optimal k-mer sizes
Adapts to your data's characteristics

3. Quality-Aware

First assembler to preserve per-base quality scores
Handles varying coverage gracefully

🚀 Quick Start (5 Minutes)

Get your first assembly running in under 5 minutes:

# 1. Install Mycelia (one-time setup)
import Pkg
Pkg.add(url="https://github.com/cjprybol/Mycelia.git")

# 2. Load the package
import Mycelia

# 3. Assemble your genome
assembly = Mycelia.assemble_genome("my_reads.fastq")

# 4. Check your results
println("Assembly complete! $(assembly.num_contigs) contigs, N50: $(assembly.n50)")

# 5. Save the assembly
Mycelia.write_fasta(assembly.contigs, "my_assembly.fasta")

That's it! Mycelia automatically:

✓ Detects your data type
✓ Selects optimal parameters
✓ Performs quality-aware assembly
✓ Validates results

🗺️ Choose Your Path

By Data Type

graph TD
    A[What type of data do you have?] --> B[FASTA only<br/>No quality scores]
    A --> C[FASTQ<br/>With quality scores]
    A --> D[Mixed<br/>Short + long reads]
    
    B --> E[K-mer Graphs<br/>→ Tutorial 1]
    C --> F[Qualmer Graphs<br/>→ Tutorial 2]
    D --> G[Hybrid Assembly<br/>→ Tutorial 3]