Technical Specifications for Next Generation Sequencing Data in Antiviral Drug Development
This guidance provides technical specifications for submitting next generation sequencing (NGS) data to FDA's Division of Antiviral Products for resistance assessments in antiviral drug development. It focuses on data submission requirements, protocols, and analysis methods to ensure proper characterization of viral resistance.
What You Need to Know? 👇
What are the key requirements for submitting NGS data to FDA’s Division of Antiviral Products?
Sponsors must submit detailed protocols, raw NGS data in fastq format, frequency tables showing amino acid substitutions ≥1%, summary tables, and comprehensive analysis methods. All submissions require thorough documentation of sequencing platforms and bioinformatics pipelines used.
Which NGS platforms does FDA accept for antiviral resistance analysis submissions?
FDA accepts data from most standard NGS platforms, provided sponsors submit appropriate platform details, sample preparation protocols, raw fastq data, and analysis methods. Early communication with the Division and mock dataset submission are recommended.
What coverage level does FDA recommend for NGS resistance analysis?
FDA recommends targeting coverage greater than 5,000 reads per sample. However, samples with lower viral loads may not achieve this coverage level, and sponsors should identify samples that didn’t reach the target coverage.
How should NGS frequency tables be formatted for FDA submission?
Frequency tables must report all amino acid substitutions differing from baseline at frequencies ≥1%. Tables should include study ID, subject ID, platform, visit, amino acid position, reference/substitution amino acids, total coverage, variant coverage, and frequency.
What are the two main approaches FDA accepts for NGS data analysis?
FDA accepts either mapping short reads to reference sequences or de novo assembly of short reads into contigs. Both approaches require detailed documentation of algorithms, parameters, variant calling methods, and quality control measures.
How should raw NGS data files be submitted to FDA?
Raw NGS data in fastq format should be submitted on secured portable hard drives following eCTD specifications. Files must be labeled with unique subject identifiers and time points (e.g., ABC123-0001.baseline.fastq). Protocols and analysis datasets go through CDER electronic gateway.
What You Need to Do 👇
Recommended Actions
- Consult with FDA Division early in the process before submitting sequencing data
- Submit a mock NGS dataset to validate data formats and processes
- Prepare detailed NGS protocol documenting sample processing and analysis procedures
- Implement proper file naming conventions and organization
- Generate required frequency and summary tables
- Document all analysis methods, algorithms and parameters thoroughly
- Ensure target coverage levels are met and documented
- Validate PCR amplification steps if used
- Prepare secured portable hard drive for raw data submission
- Set up proper data analysis pipeline with error assessment and quality control
- Implement contamination detection procedures
- Create coverage graphs and statistics for all samples
Key Considerations
Non-clinical testing
- Target coverage should be greater than 5,000 reads
- Samples not reaching target coverage should be identified
- Evidence needed that PCR amplification steps are representative of target population
- Methods for extracting, purifying and concentrating viral nucleic acids must be described
Software
- Description of analysis pipeline and algorithms used required
- Parameters and settings for mapping reads and variant calling must be documented
- Programs used for de novo assembly and annotation must be specified
Cybersecurity
- Secured portable hard drive required for raw NGS data submission
- Password protection procedures if used must be coordinated with CDER
Labelling
- Specific file naming conventions required for fastq files
- Files must be labeled with unique subject identifier and timepoint
- Resequenced samples must be distinctly labeled
Other considerations
- Raw NGS data must be in fastq format
- Frequency tables required for variants ≥1% different from baseline
- Summary tables of predominant substitutions needed
- Coverage graphs and statistics must be provided
- Contamination assessment required for unmapped reads
Relevant Guidances 🔗
- Content of Premarket Submissions for Device Software Functions
- Cybersecurity in Medical Devices: Design, Implementation, and Premarket Submissions
- Off-The-Shelf Software in Medical Devices: Documentation Requirements for Premarket Submissions
- Software Validation for Medical Device Production, Quality Systems, and Device Components
- Unique Device Identifier (UDI) Form and Content Requirements