CoV-Seq, a New Tool for SARS-CoV-2 Genome Analysis and Visualization: Development and Usability Study

Boxiang Liu, Kaibo Liu, He Zhang, Liang Zhang, Yuchen Bian, Liang HuangJournal of Medical Internet Research 2020 · DOI

TL;DR

The COVID-19 pandemic has led to an explosion of SARS-CoV-2 genomic data, overwhelming existing analysis tools. This paper introduces CoV-Seq, a novel web-based platform that simplifies and accelerates the analysis and visualization of SARS-CoV-2 genomes, making it accessible even to users without extensive bioinformatics expertise. It also provides a continuously updated database of genetic variants from public repositories, aiding global research efforts.

What Problem Does This Paper Address?

The rapid spread of COVID-19 has resulted in a massive and ever-growing volume of SARS-CoV-2 genomic sequences being deposited into public databases. Scientists need to analyze these sequences to understand the virus's evolution, identify genetic variations, and track its spread. However, this process is often ad hoc, labor-intensive, and requires specialized bioinformatics skills and programming knowledge, creating a bottleneck for many researchers. Existing tools may focus on specific aspects like gene annotation but lack a comprehensive, user-friendly solution for variant identification and visualization for this particular virus.

Methods

CoV-Seq is built using Python and JavaScript, offering both a web server and a command-line interface (CLI).

  • Data Collection: Publicly available SARS-CoV-2 sequences were gathered from major repositories: GISAID, NCBI, ENA, and CNGB. Selenium was used to automate data downloads.
  • Data Preprocessing: Incomplete genomes (less than 25,000 nucleotides) were filtered. Duplicate sequences were identified and removed by comparing accession IDs and genomic sequences.
  • Pairwise Sequence Alignment: Sequences were aligned to the SARS-CoV-2 reference genome (NC_045512.2) using MAFFT.
  • Variant Calling: A custom Python script was developed to identify single nucleotide polymorphisms (SNPs), insertions, and deletions. Variants were left-normalized using bcftools. Samples with an excessive number of variants (indicating potential sequencing errors) were removed.
  • Variant Filtering: To improve accuracy, multiallelic sites (which are more prone to sequencing errors) and variants within the poly-A tail were removed. The filtered variants were annotated using snpEff.
  • ORF Boundary Detection: A method similar to VAPiD was employed to identify open reading frame (ORF) boundaries by translating coordinates from the reference genome to the query genome.
  • Interactive Visualization: The web server, hosted on AWS, uses Flask, Jinja, and Werkzeug. It provides an interactive genome visualizer that displays genetic variants and ORF predictions. Users can zoom, pan, and hover over elements for detailed information. Results can be downloaded in VCF and JSON formats.
  • Command-Line Interface (CLI): For high-throughput processing of large datasets, a CLI is provided, allowing users to identify variants and ORF boundaries from multiple sequences efficiently.

Datasets

The primary datasets used are publicly available SARS-CoV-2 genomic sequences from:

  • Global Initiative on Sharing Avian Influenza Data (GISAID)
  • National Center for Biotechnology Information (NCBI)
  • European Nucleotide Archive (ENA)
  • China National GeneBank (CNGB)

The paper also mentions aggregating metadata associated with these sequences, including location and collection date.

Key Insights and Ideas

  • Democratizing SARS-CoV-2 Analysis: CoV-Seq significantly lowers the barrier to entry for analyzing SARS-CoV-2 genomes, empowering researchers without deep bioinformatics expertise.
  • Integrated Workflow: The tool provides a comprehensive pipeline from raw sequence input to variant and ORF annotation, visualization, and downloadable results, streamlining a complex process.
  • Importance of Data Curation: The paper highlights the necessity of robust data preprocessing steps (filtering incomplete genomes, removing duplicates and error-prone variants) to ensure the reliability of downstream analysis.
  • Value of Centralized, Updated Data: By aggregating and continuously updating a database of publicly available SARS-CoV-2 variants, CoV-Seq serves as a valuable resource for tracking viral evolution and identifying emerging patterns.

Key Results

  • CoV-Seq successfully processes custom SARS-CoV-2 sequences, automatically predicting gene boundaries and identifying genetic variants.
  • The interactive genome visualizer allows for detailed exploration of variants and ORFs, with pop-up windows providing specific information upon hovering.
  • A command-line interface is available for high-throughput processing of large batches of sequences.
  • A weekly updated database of genetic variants from all publicly accessible SARS-CoV-2 sequences is provided for download and further analysis.
  • The paper presents statistics on the geographical and chronological distribution of SARS-CoV-2 sequence submissions, illustrating the virus's global spread.

Limitations and Future Work

  • SARS-CoV-2 Specificity: CoV-Seq is currently limited to analyzing SARS-CoV-2 sequences and does not support custom reference genomes. The authors acknowledge this focus was due to the overwhelming demand for SARS-CoV-2 analysis during the pandemic.
  • Future Expansion: The authors plan to extend CoV-Seq's functionality to accept custom reference sequences and to include analysis for other viruses in future releases.

Summary updated

Abstract

We have developed CoV-Seq, an integrated web service for fast and easy analysis of custom SARS-CoV-2 sequences. The web server provides an interactive module for the analysis of custom sequences and a weekly updated database of genetic variants of all publicly accessible SARS-CoV-2 sequences. We believe CoV-Seq will help improve our understanding of the genetic underpinnings of COVID-19.
BibTeX
@article{liu2020cov,
  author = {Boxiang Liu and Kaibo Liu and He Zhang and Liang Zhang and Yuchen Bian and Liang Huang},
  title = {{CoV-Seq, a New Tool for SARS-CoV-2 Genome Analysis and Visualization: Development and Usability Study}},
  journal = {Journal of Medical Internet Research},
  year = {2020},
  doi = {10.2196/22299},
}
Figure