BreakDancer is an algorithm for high-resolution mapping of genomic structural variation. It detects various structural variants, including indels, inversions, and translocations. The algorithm was tested on simulated data, compared with other methods, and applied to an acute myeloid leukemia (AML) sample and 1,000 Genomes trio individuals. BreakDancer significantly improved the detection of small and intermediate-sized indels (10 bp to 1 Mbp) that are difficult to detect with conventional methods. It also detected a wide range of structural variants, including deletions, insertions, inversions, and translocations, with high accuracy and sensitivity. BreakDancer consists of two algorithms: BreakDancerMax and BreakDancerMini. BreakDancerMax detects large structural variants, while BreakDancerMini focuses on small indels (10-100 bp). Together, they provide accurate detection of a wide variety of structural variants. In simulations, BreakDancerMax detected 324 structural variants with a 1.48% false positive rate, including 147 shorter than 60 bp. BreakDancerMini detected 543 variants with a 7.3% FPR, including 407 shorter than 60 bp. When combined, they detected 683 variants, including 365 deletions, 290 insertions, and 21 inversions. BreakDancer was compared with other structural variant detection tools, VariationHunter and MoDIL. It showed comparable performance in detecting large fosmid deletions and had higher accuracy in detecting known deletion and insertion polymorphisms. In an AML sample, BreakDancer detected 7087 variants, including 3170 deletions, 1570 insertions, 1382 inversions, and 965 intra-chromosomal translocations. It also identified 223 putative somatic variants, including 100 deletions, 67 insertions, 22 inversions, and 34 intra-chromosomal translocations. In the 1,000 Genomes dataset, BreakDancer detected 125 deletions in NA12878, 79 of which overlapped with the Database of Genomic Variants (DGV). It also detected 246 deletions in NA19240, 123 of which overlapped with DGV. BreakDancer showed high accuracy in detecting structural variants in both family and population-based studies. It improved the specificity of somatic variant prediction by effectively eliminating inherited variants. BreakDancer's algorithms are generic and can be applied to different insert sizes and sequencing technologies. It can also be used to analyze paired-end data from mRNA sequencing to identify gene fusions and alternative splicing. The algorithm's confidence scoresBreakDancer is an algorithm for high-resolution mapping of genomic structural variation. It detects various structural variants, including indels, inversions, and translocations. The algorithm was tested on simulated data, compared with other methods, and applied to an acute myeloid leukemia (AML) sample and 1,000 Genomes trio individuals. BreakDancer significantly improved the detection of small and intermediate-sized indels (10 bp to 1 Mbp) that are difficult to detect with conventional methods. It also detected a wide range of structural variants, including deletions, insertions, inversions, and translocations, with high accuracy and sensitivity. BreakDancer consists of two algorithms: BreakDancerMax and BreakDancerMini. BreakDancerMax detects large structural variants, while BreakDancerMini focuses on small indels (10-100 bp). Together, they provide accurate detection of a wide variety of structural variants. In simulations, BreakDancerMax detected 324 structural variants with a 1.48% false positive rate, including 147 shorter than 60 bp. BreakDancerMini detected 543 variants with a 7.3% FPR, including 407 shorter than 60 bp. When combined, they detected 683 variants, including 365 deletions, 290 insertions, and 21 inversions. BreakDancer was compared with other structural variant detection tools, VariationHunter and MoDIL. It showed comparable performance in detecting large fosmid deletions and had higher accuracy in detecting known deletion and insertion polymorphisms. In an AML sample, BreakDancer detected 7087 variants, including 3170 deletions, 1570 insertions, 1382 inversions, and 965 intra-chromosomal translocations. It also identified 223 putative somatic variants, including 100 deletions, 67 insertions, 22 inversions, and 34 intra-chromosomal translocations. In the 1,000 Genomes dataset, BreakDancer detected 125 deletions in NA12878, 79 of which overlapped with the Database of Genomic Variants (DGV). It also detected 246 deletions in NA19240, 123 of which overlapped with DGV. BreakDancer showed high accuracy in detecting structural variants in both family and population-based studies. It improved the specificity of somatic variant prediction by effectively eliminating inherited variants. BreakDancer's algorithms are generic and can be applied to different insert sizes and sequencing technologies. It can also be used to analyze paired-end data from mRNA sequencing to identify gene fusions and alternative splicing. The algorithm's confidence scores