Document for SOAPsplice users. Here we introduce the format of the output files. There are three kinds of output files: 1.One-segment alignment result: *.out 2.Two-segment alignment result: *.2Segs 3.Junction file: *.junc Format of one-segment alignment file (file name is "*.out"): 1)index: num "i" means it's ith read; 2)id: the id of the read; 3)seq: sequence of mapped bases. The read will be converted to the reverse complement of the sequence if mapped on the reverse chain of reference; 4)qual: quality of sequence. Corresponding to sequence, to be consistent with seq, it will be converted too if mapped on reverse chain; 5)number of hits: number of equal best hits. The reads with no hits will be ignored; 6)length: length of mapped bases; 7)+/-: alignment on the direct(+) or reverse(-) chain of the reference; 8)chr: id of reference sequence; 9)location: location of first bp on the reference, counted from 1; 10)types: type of hits. "0": exact match. "1~3 RefAllele->OffsetQueryAlleleQual": number of mismatches, followed by detailed mutation sites and switch of allele types. Offset is relative to the initial location on reference. 'OffsetAlleleQual': offset, allele, and quality. "1INDEL", "2INDEL": alignment with 1 or 2 indels. Example: "2 A->10T30 C->13A32" means there are two mismatches, one on location+10 of the reference, and the other on location+13 of reference. The alleles on reference are A and C respectively, while query alleles and their quality are T,30 and A,32. Format of two-segment alignment file (file name is "*.2Segs"): This file has the similar format with one-segment alignment file, except that some column has two information separated by a comma, in such columns, the left information is for the first segment, while the right information is for the second segment. 1)index: num "i" means it's ith read; 2)id: the id of the read; 3)hitOne,hitTwo: number of equal best hits for both segments; 4)lengthOne,lengthTwo: the length for both segments; 5)+/-: alignment on the direct(+) or reverse(-) chain of the reference; 6)direction: the chain that the junction reported by this read is on, "fwd" means it's on the direct chain, while "rev" means it's on the reverse chain; 7)chrOne,chrTwo: id of reference sequence that both segments align to; 8)locationOne,locationTwo: location of first bp on the reference for both segments, counted from 1; 9)mismatchOne,mismatchTwo: the number of mismatches for both segments, at most 1 for each segment; 10)mismatchInfo: the same as last column in one-segment alignment file. Format of junction file (file name is "*.junc"): 1)chr: id of reference sequence that the junction comes from; 2)site1: left site of the junction site, one previous the left bound of the intron; 3)site2: right site of the junction site, one after the right round of the intron; 4)direction: the chain that the intron is on, "fwd" means it's on the direct chain, while "rev" means it's on the reverse chain. 5)number of reads supportting this junction.