v2.0.0 – Reforged Core

🚀 SvPhaser v0.2 — “Fork & Phase”
Highlights

Massive speed-up – true parallelism via new _workers.py (one process per chromosome).

Clean type layer – central types.py (NamedTuples + dataclasses) for safer code & IDE happiness.

INFO-rich outputs – phased VCF now carries HP_GT, HP_GQ, and binned HP_GQBIN labels.

Smarter CLI – clearer help, sensible defaults (--threads auto-detects cores), and colorised logs.

Robust I/O – header cloning no longer depends on VCF.copy(); works with any cyvcf2 build.

Docs & badges – polished README, Mermaid methodology diagram, sample result images & stats.

CHANGELOG – all changes now tracked; semantic versioning begins here.

Thanks

Developed by Team 5 (BioAI Hackathon) — Pranjul Mishra & Sachin Gadakh
Feedback and PRs welcome!

Foundation Stone

🚀 SvPhaser v1.0.0 – Foundation Stone

The first public release of SvPhaser, a high-performance, open-source tool for phasing structural variants (SVs) using phased long-read alignments.

🧬 Problem Solved:
Structural variant phasing remains a major gap in genomics, limiting allele-specific analysis and trio validation. Most existing SV callers detect variants but do not assign haplotypes, making compound heterozygosity and inheritance tracking difficult.

🛠️ Core Features in This Release:
CLI tool to phase pre-detected SVs using phased BAM files (HP tags)

Majority-based haplotype assignment (≥80% read support threshold)

Confidence classification: High, Moderate, Low

SV-type-aware phasing logic (DEL, INS, INV, etc.)

Output formats:

Per-chromosome and merged CSVs

Phased VCF with GT:DP fields

📊 Statistical Enhancements:
Confidence score based on HP1 vs HP2 read imbalance

SV density plots and support histograms

🔬 Data Used for Initial Validation:
Real trio long-read datasets from 1000 Genomes (Puerto Rican, Han-Chinese, Yoruban)

🌱 This Release Marks:
The foundation of a lightweight yet powerful SV phasing solution

The beginning of a future-ready tool designed to integrate with Hi-C, Strand-seq, and trio-aware pipelines