Simrank: Rapid and sensitive general-purpose k-mer search tool

BMC Ecology

Table 1 Simrank database binary file structure and storage requirements.

File Segment	File Element	Storage Requirement (bytes)
1	F, string ID field size	10
2	K, k-mer length	10
3	N, string count	10
4	string ID array	FN
5	offset arrays^a
6	k-mer array^b	Kk
7	offsets index array^c	4k
8	offsets lengths array^d	4k
9	unique k-mers per string array^e	4N
10	k, unique k-mer count	10
11	file position of segment 6	10

^aEach k-mer generates a vector of string indices, encoded as an integer array of offsets required to "visit" each string index containing the k-mer. k is the count of unique k-mers, and s_i is the count of strings containing the i^th k-mer. Each offset is stored as a 4-byte integer.
^bLexically sorted ASCII text strings of each unique k-mer stored as one byte per character.
^c4-byte integer list of file positions for the start of each k-mer's list of offsets.
^d4-byte integer list of the byte length of each k-mer's list of offsets.
^e4-byte integer list of the count of unique k-mers in each string.

ISSN: 1472-6785