domains / families / science / long-read-qc

long-read-qc

separate

Assesses Oxford Nanopore and PacBio long-read quality with NanoPlot, cramino, NanoComp, pycoQC/toulligQC, and seqkit, and filters reads with chopper/Filtlong for the downstream goal. Covers why read-only Qscore is an uncalibrated posterior (real accuracy needs a reference BAM), why the sequencing_summary.txt is required for run-health metrics, intent-conditioned filtering (preserve long reads and small replicons for assembly, filter almost nothing for variant calling), the chimera/internal-adapter trap that fabricates SVs, and PacBio rq-based HiFi QC. Use when judging a long-read run, computing read N50 or percent identity, filtering reads before assembly or variant calling, comparing barcodes/runs, or reading run-health red flags.

When to use

• judging a long-read run
• computing read N50 or percent identity
• filtering reads before assembly or variant calling
• comparing barcodes/runs
• reading run-health red flags
• PacBio HiFi QC

When not to use

• short-read QC

Skill page

Source