BWA is a fast and accurate short read aligner based on the Burrows-Wheeler Transform (BWT), designed for aligning short sequencing reads against a large reference genome. It supports both base space and color space reads, and outputs alignments in the SAM format. BWA is approximately 10–20 times faster than MAQ while maintaining similar accuracy. It efficiently handles gapped alignments and allows for mismatches and gaps, making it suitable for longer reads with frequent indels. BWA uses a backward search approach with BWT to mimic top-down traversal on the prefix trie of the genome, enabling efficient alignment with minimal memory usage. It also incorporates a bounded traversal/backtracking method to limit search space and improve efficiency. BWA supports paired-end mapping and generates mapping quality scores, which are crucial for downstream analyses. It is memory-efficient, using about 2.3 GB for single-end mapping and 3 GB for paired-end, and supports multi-threading. BWA is compared with other aligners like MAQ, SOAPv2, and Bowtie on simulated and real data, showing superior performance in terms of speed and accuracy. It is particularly effective for short reads and can handle long reads with Smith–Waterman alignment. BWA is also capable of aligning SOLiD reads by converting the reference genome to a dinucleotide color sequence. The algorithm uses dynamic programming to decode color reads into nucleotide sequences, ensuring accurate alignment. BWA's performance is evaluated on both simulated and real data, demonstrating its effectiveness in aligning reads with high accuracy and efficiency.BWA is a fast and accurate short read aligner based on the Burrows-Wheeler Transform (BWT), designed for aligning short sequencing reads against a large reference genome. It supports both base space and color space reads, and outputs alignments in the SAM format. BWA is approximately 10–20 times faster than MAQ while maintaining similar accuracy. It efficiently handles gapped alignments and allows for mismatches and gaps, making it suitable for longer reads with frequent indels. BWA uses a backward search approach with BWT to mimic top-down traversal on the prefix trie of the genome, enabling efficient alignment with minimal memory usage. It also incorporates a bounded traversal/backtracking method to limit search space and improve efficiency. BWA supports paired-end mapping and generates mapping quality scores, which are crucial for downstream analyses. It is memory-efficient, using about 2.3 GB for single-end mapping and 3 GB for paired-end, and supports multi-threading. BWA is compared with other aligners like MAQ, SOAPv2, and Bowtie on simulated and real data, showing superior performance in terms of speed and accuracy. It is particularly effective for short reads and can handle long reads with Smith–Waterman alignment. BWA is also capable of aligning SOLiD reads by converting the reference genome to a dinucleotide color sequence. The algorithm uses dynamic programming to decode color reads into nucleotide sequences, ensuring accurate alignment. BWA's performance is evaluated on both simulated and real data, demonstrating its effectiveness in aligning reads with high accuracy and efficiency.