Table 1 Algorithms for error correction in Ion Torrent PGM data.
From: Comparison of error correction algorithms for Ion Torrent PGM data: application to hepatitis B virus
Method | Algorithm | Comment | Quality score | Input file | Target error type | Ref. |
|---|---|---|---|---|---|---|
Fiona | Suffix array/tree | Use a suffix tree to detect and correct substitution and indel errors, and use edit distance comparisons to enhance overlap detection of indel errors. | Not used | fasta/fastq | Substitution Deletion/Insertion | |
Pollux | k-spectrum | Divide all across reads into k-mer lengths, count the observed k-mer numbers, and generate k-mer depth profiles to correct the k-mer profiles. Compare the adjacent k-mers and identify discontinuities to find error locations and evaluate correctness. | Not specifically used | fastq | Substitution Deletion/Insertion | |
Blue | k-spectrum | Tile reads to reduce the k-mer spectrum, distinguish k-mers from organisms or containing sequencing error reads, and choose between alternative replacement k-mers and a k-mer spectrum trust threshold to correct the reads. | Not used | fasta/fastq | Substitution Deletion/Insertion | |
Karect | MSA | Take each read r as a reference and perform multiple alignments by selecting optimized reads similar to r; represent graph reads; and compute graph edge weights and construct corrected reads. | Not used | fasta/fastq | Substitution Deletion/Insertion | |
Coral | MSA | Compute initial read overlaps with hash tables to the k-mer length, form multiple alignments of the reads and rely on quality scores to distinguish and correct erroneous bases. | Used | fasta/fastq | Substitution Deletion/Insertion |
