Data analysis: Difference between revisions
Expand Nanopore Wiki source page with structured educational content |
Use topic-specific Wikimedia Commons illustrations |
||
| Line 2: | Line 2: | ||
<translate> | <translate> | ||
<!--T:1--> | <!--T:1--> | ||
[[File: | [[File:Detection pics scipy signal find peaks.svg|thumb|right|320px|Peak or event detection concepts support nanopore signal analysis. Source: Wikimedia Commons, CC BY-SA 4.0.]] | ||
'''Data analysis''' is part of the Nanopore Wiki learning path. Nanopore data analysis extracts reliable events, features, and conclusions from raw current traces or sequencing output. | '''Data analysis''' is part of the Nanopore Wiki learning path. Nanopore data analysis extracts reliable events, features, and conclusions from raw current traces or sequencing output. | ||
Revision as of 10:57, 30 June 2026
Data analysis is part of the Nanopore Wiki learning path. Nanopore data analysis extracts reliable events, features, and conclusions from raw current traces or sequencing output.
Core concepts
- For sensing, raw data are often current-versus-time traces.
- For sequencing, data may include POD5, FAST5, FASTQ, BAM, and reports.
- Analysis should preserve true event features while controlling noise, drift, and artifacts.
- Quality control should be performed before and after feature extraction.
Analysis workflow
| Step | Description |
|---|---|
| 1 | Organize metadata and raw files. |
| 2 | Assess baseline, noise, and drift. |
| 3 | Filter or detrend only with documented parameters. |
| 4 | Detect events and extract features such as blockade depth and dwell time. |
| 5 | Visualize both representative traces and population statistics. |
Typical applications
- Single-molecule sensing.
- Sequencing quality control.
- Method comparison and troubleshooting.
Common pitfalls
- Showing only attractive example traces.
- Changing thresholds until expected results appear.
- Publishing feature tables without raw-data provenance.