Coral is an error correction algorithm for correcting reads from DNA sequencing platforms such as the Illumina Genome Analyzer or HiSeq platforms or Roche/454 Genome Sequencer.

Sources and example files:

The above read set is simulated data. It is generated from the Escherichia Coli genome K-12 substrain MG1655 (NC_000913). The error rate of the reads is about 3 % and the coverage is 12. The set contains 559637 reads.

Old versions:


Is it possible to use Coral to correct errors in 454 reads utilizing Illumina reads?

The latest (1.4) version of Coral can correct e.g. 454 reads using Illumina reads. Note that Coral reads all its input reads from a single file (so one can only use -fq or -fs once and not both at the same time). The trick is to put all your reads into the same file so that the Illumina reads come first and then the 454 reads. Then if you have <x> Illumina reads, you can use the options -i <x> -j <x> to tell Coral to index only the first <x> reads but not to correct the first <x> reads. This will make Coral to take one 454 read at a time, fetch (Illumina) reads to align from the index and then correct the 454 read. Optionally, if you want to also correct the Illumina reads you can just use only the option -i <x>.