# Sequencing depth and coverage

Source: https://onco.cc/terms/sequencing-depth-and-coverage/  
OnCo record `sequencing-depth-and-coverage` (Term). Data CC BY-NC 4.0, attribute "Data from OnCo (onco.cc)"; commercial use needs a licence.

## TL;DR

Depth counts sequence observations at a position; breadth describes how much of a target meets a stated depth.

## Summary

Depth is the number of aligned reads covering a base after stated filters; 40x means forty counted observations under those rules; depth without its counting rule is incomplete.

## Fields

- Kind: Term
- Last checked: 2026-11-15

## Notes

- Depth is the number of aligned reads covering a base after stated filters; 40x means forty counted observations under those rules; depth without its counting rule is incomplete.
- Mean depth averages positions; coverage breadth is the fraction of a named target (denominator: target bases, not reads or cells) meeting a stated minimum depth; neither shows which important positions were missed.
- Fictional report: mean 100x, 900 of 1,000 target bases at or above an illustrative 20x criterion gives 90% breadth, while the position of interest has 4 usable observations; 20x is not a clinical cutoff. The mean is a supplied report value: no full per-base profile is given from which to recompute it. Four read observations also do not prove four independent original molecules.
- Read depth differs from molecular depth: overlapping pair ends and amplification copies inflate reads without adding independent fragments; HTSlib defaults are software definitions, not clinical validation criteria.
- More sequencing cannot recover a region absent from the library or create independent molecules from duplicates; library complexity, input, preservation, capture and alignment limit usable evidence.
- A negative report covers tested regions, detectable change classes and validated sensitivity; identical depths can support different detection limits; depth does not establish origin, function or actionability.
- A narrow panel can be deep and a genome broad but thinner per position; territory asks where, depth asks how many.
- Use an existing alignment file to calculate depth without consuming more tissue; recovering a more diverse library may require another aliquot. Retain reference regions, per-position counts, summary depth, breadth criterion and molecule-count rules. Capture bias, duplicate inflation and damaged input require controls and intended-use sensitivity/reproducibility validation; a summary criterion is separate from a detection limit.
- In a separate fictional target, four usable aligned reads overlap position P, only two overlap Q, and none overlap R even though R was selected. Their local depths are 4x, 2x and 0x. R has no usable observations in this example; a position outside the assay design was not examined. These values establish neither a clinical threshold nor a variant-detection probability.

## Sources

- Guidelines for Validation of Next-Generation Sequencing-Based Oncology Panels: A Joint Consensus Recommendation of the Association for Molecular Pathology and College of American Pathologists.: https://pubmed.ncbi.nlm.nih.gov/28341590/
- samtools depth counting rules: https://www.htslib.org/doc/samtools-depth.html
- samtools coverage summary rules: https://www.htslib.org/doc/samtools-coverage.html

## Connected records

- terms: [Next-generation sequencing (NGS)](https://onco.cc/terms/ngs/)

---
JSON: https://onco.cc/api/v1/entities/sequencing-depth-and-coverage.json