Introduction #
SeqSMART follows a structured and transparent workflow that ensures every genomic case is analyzed with precision, reproducibility, and traceability.
The workflow reflects the real-world process of variant interpretation — from initial data upload to the final generation of a comprehensive analysis report.
This document provides an overview of each phase within the SeqSMART pipeline and the key actions performed in every step.
1. Case Creation #
The workflow begins with the creation of a new case in the Case Manager module.
At this stage, users define the analysis parameters and upload relevant input data.
Inputs may include:
- VCF files: Primary variant data from sequencing.
- Phenotype data: HPO terms or clinical features associated with the subject.
- Pedigree data: Family relationships for segregation analysis.
- Metadata: Sample identifiers, sequencing type, and reference genome version.
Once a case is created, SeqSMART assigns a unique case ID and logs it into the database with all associated files and metadata.
2. Automated Variant Annotation #
After a case is initialized, SeqSMART automatically performs variant annotation.
This step enriches each variant with detailed biological and clinical information from multiple integrated sources.
Annotation includes:
- Gene and transcript mapping (based on reference genome and selected transcript).
- Functional consequence (missense, nonsense, splice, etc.).
- Population frequency data (gnomAD, ExAC, 1000 Genomes).
- ClinVar and OMIM associations.
- Computational predictions (SIFT, PolyPhen, CADD, etc.).
- Conservation and domain context.
All annotations are stored in the case database for downstream analysis and reporting.
3. Automated ACMG Classification #
The annotated variants are then passed through SeqSMART’s Variant Interpretation Engine, which automatically evaluates them according to the ACMG/AMP guidelines.
Each criterion (PVS1, PS, PM, PP, BS, BP categories) is tested systematically using available evidence.
SeqSMART applies a weighted logic framework that integrates:
- Population data (benign evidence)
- Functional predictions (supporting/moderate evidence)
- Segregation or inheritance data
- Pathogenic variant databases
- Phenotypic matching via HPO terms
The system then assigns an initial classification (Pathogenic, Likely Pathogenic, VUS, Likely Benign, or Benign) and provides a complete breakdown of the supporting rules and evidence.
Each ACMG rule is fully traceable and can be reviewed, modified, or re-evaluated by analysts.
4. Analyst Review #
Once automated classification is completed, the case progresses to the Analyst Review stage.
At this level, expert users review all system-generated classifications, confirm or adjust evidence, and add manual annotations if necessary.
Analyst actions may include:
- Reviewing ACMG rule justifications and underlying data.
- Adding literature-based or experimental evidence not captured automatically.
- Reclassifying variants based on expert interpretation.
- Adding clinical comments or summary notes.
SeqSMART logs every analyst action for audit purposes and records any classification overrides or rationale.
5. Checker Validation #
After analyst review, the case enters the Checker Validation stage for secondary evaluation.
A designated checker (or senior reviewer) validates the analysis to ensure consistency, accuracy, and completeness before final approval.
Validation tasks include:
- Confirming all evidence aligns with institutional standards.
- Ensuring report comments and classifications are scientifically sound.
- Approving or rejecting the case for final reporting.
This two-layer review structure enhances quality assurance and prevents oversight in clinical or research reporting.
6. Report Generation #
Once approved, SeqSMART automatically compiles the final report, integrating all validated findings, comments, and supporting evidence.
Reports can be generated in customizable templates (e.g., institutional or laboratory-specific formats).
Each report includes:
- Case information and metadata
- Summary of analysis steps and QC metrics
- Final variant classifications with ACMG rule breakdown
- Phenotype and inheritance summary
- Analyst and checker comments
- References and supporting databases
Reports are available for download in multiple formats such as PDF, JSON, or CSV, depending on institutional preferences.
7. Archiving and Traceability #
After report generation, the case is archived in SeqSMART’s secure data repository.
All case components — input files, annotations, decisions, and reports — remain linked and accessible through their unique identifiers.
SeqSMART maintains a complete audit trail, enabling users to trace every decision, update, or version change throughout the case’s lifecycle.
Archived cases can be retrieved or reanalyzed at any time as new evidence or database updates become available.
The SeqSMART workflow ensures that each variant analysis follows a standardized, multi-stage process — from raw data intake to expert-reviewed, traceable reporting.
This structured pipeline provides transparency, reproducibility, and compliance for genomic professionals who rely on SeqSMART for precision and reliability.
SeqSMART Workflow Principle:
Automate where possible, review where essential, and document everything.