Restoring flowcell type and basecaller configuration from FASTQ files of nanopore sequencing data

Area of Science:

• Genomics
• Bioinformatics
• Computational Biology

Background:

• Nanopore sequencing generates vast public datasets, crucial for genomic research.
• Key metadata (flowcell type, basecaller configuration) is missing in ~85% of datasets, hindering analysis.
• Accurate metadata is vital for bioinformatics algorithms and reproducible results.

Purpose of the Study:

• To develop LongBow, a tool for inferring nanopore sequencing flowcell type and basecaller configuration.
• To improve the usability and reproducibility of public nanopore sequencing data.
• To enhance variant calling and lineage assignment in genomic studies.

Main Methods:

• LongBow analyzes base quality value patterns within FASTQ files to predict metadata.
• The tool was validated on diverse in-house and public datasets.
• A LongBow-based pipeline was applied to COVID-19 Genomics UK (COG-UK) data.

Main Results:

• LongBow achieved high accuracy: 95.33% on in-house data and 91.45% on public FASTQ files.
• Reanalysis of COG-UK data confirmed LongBow's necessity for variant reproduction.
• The LongBow pipeline identified more functionally important variants and improved lineage assignment accuracy.

Conclusions:

• LongBow effectively infers essential nanopore sequencing metadata from FASTQ files.
• The tool significantly boosts the utility and reproducibility of public nanopore data.
• LongBow is critical for advancing genomic research and accurate biological interpretation.