Date of Award


Degree Name

Doctor of Philosophy


Computer Science

First Advisor

Dr. Elise de Doncker

Second Advisor

Dr. Alvis Fong

Third Advisor

Dr. Todd Barkman

Fourth Advisor

Dr. Nancy Deng


motif finding, approximate algorithms, ChIP-Seq, DNA sequences


Motif discovery is the problem of finding common substrings within a set of biological strings. Therefore it can be applied to finding Transcription Factor Binding Sites (TFBS) that have common patterns (motifs). A transcription factor molecule can bind to multiple binding sites in the promoter region of different genes to make these genes co-regulating. The Planted (l, d) Motif Problem (PMP) is a classic version of motif discovery where l is the motif length and d represents the maximum allowed mutation distance. The quorum Planted (l, d, q) Motif Problem (qPMP) is a version of PMP where the motif of length l occurs in at least q percent of the sequences with up to d mismatches. In this thesis we develop the Strong Motif Finder (SMF) and quorum Strong Motif Finder (qSMF) algorithms and evaluate their performance.

The Strong Motif Finder (SMF) returns a list of its highest ranked (strongest) motifs. The performance of SMF is compared with the APMotif and MEME algorithms with respect to execution time and prediction accuracy. Several performance metrics are used at both the nucleotide and the site level. The algorithms are tested on simulated datasets. The time comparisons show that SMF is faster than the APMotif and the MEME (ANR) and similar in speed to the MEME (ZOOPS). The MEME algorithm with choice OOPS is the fastest but is not practical if no prior knowledge is available. The prediction accuracy results reveal that the SMF outperforms the APMotif, and performs at the level of the best prediction accuracy of the MEME (with OOPS choice), notwithstanding that the SMF is not given a-priori information. In addition, the SMF is tested on real DNA datasets of orthologous regularity regions from multiple species, without using their related phylogenetic tree. The experiments indicate that the SMF results agree with published motifs.

The quorum Strong Motif Finder (qSMF) returns a list of highest ranked (strongest) motifs occurring in at least q percent of the data sequences. The algorithm is tested on ChIPSeq (large) data that was sampled using the SamSelect algorithm. In comparison with the FMotif algorithm, the experimental results show that qSMF is faster and returns predicted motifs similar to results in the literature and to motifs discovered by the ENCODE project tool which uses the established motif finding algorithms of AlignACE, MEME, MDscan, Trawler, and Weeder.

In order to determine the strength or the significance of the predicted motifs, a scoring function, the Motif Strength Score (MSS), is proposed for ranking the discovered motifs in both algorithms. In future work, this score can be combined with other statistical scores, such as the complexity score, P-value and information content, to better determine the motif significance.

Access Setting

Dissertation-Open Access