Second, SEQMINER implements novel data structures for storing bioinformatics databases to speed up annotations and query. To support efficient region‐based variant annotations, we used a segment tree data structure, enabling queries of O(log(N))time complexity and construction of O(N log(N))time and space complexity where N is the number of regions. Specifically, we constructed a red‐black tree and stored every region (start position, end position, and region names) in a tree leaf. These regions are ordered internally by their start positions and then end positions. As the tree is balanced, querying genomic variants requires at most log2(N) comparisons. This data structure has high performance in practice, and it enables the online annotation of range‐based databases such as transcription factor binding sites.
