PairMotif: A New Pattern-Driven Algorithm for Planted (l, d) DNA Motif Search

Abstract

Motif search is a fundamental problem in bioinformatics with an important application in locating transcription factor binding sites (TFBSs) in DNA sequences. The exact algorithms can report all (l, d) motifs and find the best one under a specific objective function. However, it is still a challenging task to identify weak motifs, since either a large amount of memory or execution time is required by current exact algorithms. A new exact algorithm, PairMotif, is proposed for planted (l, d) motif search (PMS) in this paper. To effectively reduce both candidate motifs and scanned l-mers, multiple pairs of l-mers with relatively large distances are selected from input sequences to restrict the search space. Comparisons with several recently proposed algorithms show that PairMotif requires less storage space and runs faster on most PMS instances. Particularly, among the algorithms compared, only PairMotif can solve the weak instance (27, 9) within 10 hours. Moreover, the performance of PairMotif is stable over the sequence length, which allows it to identify motifs in longer sequences. For the real biological data, experimental results demonstrate the validity of the proposed algorithm.

Ứng Dụng Giải Thuật Song Song Trên Hệ Thống Cpu-Gpu Cho Bài Toán Tìm Kiếm Motif

“…Chương trình xử lý song song được lập trình trên hai bộ công cụ là CUDA và OpenCL để tiến hành so sánh hiệu suất của chúng. Chương trình sẽ sử dụng bộ dữ liệu từ [7] để tiến hành đánh giá kết quả của thuật toán Pattern Branching xử lý song song và xử lý tuần tự. Các tập trình tự thử nghiệm gồm 20 trình tự với chiều dài mỗi trình tự là 600, đây là số lượng trình tự và chiều dài được đặt ra trong bài toán thách thức tìm motif [8], chiều dài và số đột biến của các motif cần tìm khác nhau trong từng trường hợp.…”

Reference sequence selection for motif searches

“…For the exact algorithms proposed in earlier years, such as WINNOWER [3], their search space is composed of (n -I + 1 Y possible alignments of motif instances. Tn recent years, the exact algorithms verifY all the I-length patterns in the O(III ' ) search space, and they are called the pattern-driven PMS algorithms [5][6][7][8][9][10][11][12][13].…”

“…Specifically, they generate candidate motifs by using all possible h-tuple T = (x), Xz ... Xh) composed of h 1-length strings coming from h distinct reference sequences. In existing pattern-driven PMS algorithms, h is 1 for PMSP [5] and PMSPrune [6]; h is 2 for PairMotif [7], qPMS7 [8] and TravStrR [9]; h is 3 for iTriplet [10] and PMS5 [11]; for PMS8 [12] and qPMS9 [13], h is greater than or equal to 3 and self adaptive in dealing with different PMS problem instances. Moreover, these algorithms use k = t -q + h reference sequences to generate candidate motifs, ensuring that there exists at least one h-tuple T so that each I-length string in T is a moti f instance.…”

Load Balancing for Parallel Motif Discoveries

“…While past research works have focused on improving the algorithms [4], [5], [6], [7], [8], [9], [10], and on designing hardware accelerators [11], [12], [13], [14], [15], for motif discovery, none of them focused on the load balancing issues between the master and the worker processes in the parallel execution. To the best of our knowledge this is the first paper that explored and evaluated all parameters of the parallel motif discovery problem to find the right balance for workload distributions.…”