What it does

BBDuk was developed to combine most common data-quality-related trimming, filtering, and masking operations into a single high-performance tool. It is capable of quality-trimming and filtering, adapter-trimming, contaminant-filtering via kmer matching, sequence masking, GC-filtering, length filtering, entropy-filtering, format conversion, histogram generation, subsampling, quality-score recalibration, kmer cardinality estimation, and various other operations in a single pass. Specifically, any combination of operations is possible in a single pass with the exception of kmer-based operations (kmer trimming, kmer masking, or kmer filtering). At most 1 kmer-based operation can be done in a single pass.

Options

• Reference - if a reference is specified, BBDuk will operate on kmers in one of 4 modes: right-trimming, left-trimming, masking, or filtering. The default is filtering - any read matching a reference kmer will be discarded.
• Trim reads to remove bases matching reference kmers - When trimming to the right, once a reference kmer is matched in a read, that kmer and all the bases to the right will be trimmed, leaving only the bases to the left. When trimming to the left, trimming will be done to the left instead.

Outputs

• Unmatched - All the reads that pass all filtering criteria. Reads will be at least as long as Minimum read length after any trimming operations and reads will not match any reference kmer if kmer-filtering is being performed. A read’s average quality will be at least as high as the specified Minimum average quality.
• Matched - Reads failing any filter criteria (such as matching a reference kmer). By default, if either read in a pair fails, both will be included in Matched.
• Single - Singleton reads whose mate was trimmed shorter than the value of Minimum read length.