MKMuhammed Oguzhan Kulekci

Muhammed Oguzhan Kulekci

*This is an algorithmically generated author record based on publicly available information.

Current Organizations / Institutes

The organizations/institutes from the scholar's latest article

Previous Organizations / Institutes

The organizations / institutes from the scholar's other articles in the past ten years


Muhammed Oguzhan Kulekci has a total of 57 co-authors. Here are the top co-authors by number of shared publications:


Latest Publications

  • Open Access

    An extended de Bruijn graph for feature engineering over biological sequential data

    by Mert Onur Cakiroglu, Hasan Kurban, Parichit Sharma, M Oguzhan Kulekci, Elham Khorasani Buxton, Maryam Raeeszadeh-Sarmazdeh, Mehmet M Dalkilic

    In this study, we introduce a novel de Bruijn graph (dBG) based framework for feature engineering in biological sequential data such as proteins. This framework simplifies feature extraction by dynamically generating high-quality, interpretable features for traditional AI (TAI) algorithms. Our framework accounts for amino acid substitutions by efficiently adjusting the edge weights in the dBG using a secondary trie structure. We extract motifs from the dBG by traversing the heavy edges, and then incorporate alignment algorithms like BLAST and Smith–Waterman to generate features for TAI algorithms. Empirical validation on TIMP (tissue inhibitors of matrix metalloproteinase) data demonstrates significant accuracy improvements over a robust baseline, state-of-the-art PLM models, and those from the popular GLAM2 tool. Furthermore, our framework successfully identified Glycine and Arginine-rich motifs with high coverage, highlighting it is potential in general pattern discovery.


  • Open Access

    A Survey on Shortest Unique Substring Queries

    by Paniz Abedin, M. Oğuzhan Külekci, Shama V. Thankachan

    The shortest unique substring (SUS) problem is an active line of research in the field of string algorithms and has several applications in bioinformatics and information retrieval. The initial version of the problem was proposed by Pei et al. [ICDE’13]. Over the years, many variants and extensions have been pursued, which include positional-SUS, interval-SUS, approximate-SUS, palindromic-SUS, range-SUS, etc. In this article, we highlight some of the key results and summarize the recent developments in this area.


  • The order-preserving pattern matching problem in practice

    by Domenico Cantone, Simone Faro, M. Oğuzhan Külekci


  • Open Access

    Enumerative Data Compression with Non-Uniquely Decodable Codes

    by M. Oğuzhan Külekci, Yasin Öztürk, Elif Altunok, Can Altıniğne

    Non-uniquely decodable codes can be defined as the codes that cannot be uniquely decoded without additional disambiguation information. These are mainly the class of non-prefix-free codes, where a codeword can be a prefix of other(s), and thus, the codeword boundary information is essential for correct decoding. Although the codeword bit stream consumes significantly less space when compared to prefix--free codes, the additional disambiguation information makes it difficult to catch the performance of prefix-free codes in total. Previous studies considered compression with non-prefix-free codes by integrating rank/select dictionaries or wavelet trees to mark the code-word boundaries. In this study we focus on another dimension with a block--wise enumeration scheme that improves the compression ratios of the previous studies significantly. Experiments conducted on a known corpus showed that the proposed scheme successfully represents a source within its entropy, even performing better than the Huffman and arithmetic coding in some cases. The non-uniquely decodable codes also provides an intrinsic security feature due to lack of unique-decodability. We investigate this dimension as an opportunity to provide compressed data security without (or with less) encryption, and discuss various possible practical advantages supported by such codes.

  • Privacy–Preserving Text Similarity via Non-Prefix-Free Codes

    by M. Oğuzhan Külekci, Ismail Habib, Amir Aghabaiglou


  • Open Access

    Applications of Non-Uniquely Decodable Codes to Privacy-Preserving High-Entropy Data Representation

    by Muhammed Oğuzhan Külekci, Yasin Öztürk

    Non-uniquely-decodable (non-UD) codes can be defined as the codes that cannot be uniquely decoded without additional disambiguation information. These are mainly the class of non–prefix–free codes, where a code-word can be a prefix of other(s), and thus, the code-word boundary information is essential for correct decoding. Due to their inherent unique decodability problem, such non-UD codes have not received much attention except a few studies, in which using compressed data structures to represent the disambiguation information efficiently had been previously proposed. It had been shown before that the compression ratio can get quite close to Huffman/Arithmetic codes with an additional capability of providing direct access in compressed data, which is a missing feature in the regular Huffman codes. In this study we investigate non-UD codes in another dimension addressing the privacy of the high-entropy data. We particularly focus on such massive volumes, where typical examples are encoded video or similar multimedia files. Representation of such a volume with non–UD coding creates two elements as the disambiguation information and the payload, where decoding the original data from these elements becomes hard when one of them is missing. We make use of this observation for privacy concerns. and study the space consumption as well as the hardness of that decoding. We conclude that non-uniquely-decodable codes can be an alternative to selective encryption schemes that aim to secure only part of the data when data is huge. We provide a freely available software implementation of the proposed scheme as well.

  • Optimizing Packed String Matching on AVX2 Platform

    by Mehmet Akif Aydoğmuş, M. Oğuzhan Külekci

  • Open Access

    Sketching Algorithms for Genomic Data Analysis and Querying in a Secure Enclave

    by Can Kockan, Kaiyuan Zhu, Natnatee Dokmai, Nikolai Karpov, Oguzhan Kulekci, David Woodruff, Cenk Sahinalp

    Current practices in collaborative genomic data analysis (e.g. PCAWG) necessitate all involved parties to exchange individual patient data and perform all analysis locally, or use a trusted server for maintaining all data to perform analysis in a single site (e.g. the Cancer Genome Collaboratory). Since both approaches involve sharing genomic sequence data - which is typically not feasible due to privacy issues, collaborative data analysis remains to be a rarity in genomic medicine. In order to facilitate efficient and effective collaborative or remote genomic computation we introduce SkSES (Sketching algorithms for Secure Enclave based genomic data analysiS), a computational framework for performing data analysis and querying on multiple, individually encrypted genomes from several institutions in an untrusted cloud environment. Unlike other techniques for secure/privacy preserving genomic data analysis, which typically rely on sophisticated cryptographic techniques with prohibitively large computational overheads, SkSES utilizes the secure enclaves supported by current generation microprocessor architectures such as Intel's SGX. The key conceptual contribution of SkSES is its use of sketching data structures that can fit in the limited memory available in a secure enclave. While streaming/sketching algorithms have been developed for many applications in computer science, their feasibility in genomics has remained largely unexplored. On the other hand, even though privacy and security issues are becoming critical in genomic medicine, available cryptographic techniques based on, e.g. homomorphic encryption or garbled circuits, fail to address the performance demands of this rapidly growing field. The alternative offered by Intel's SGX, a combination of hardware and software solutions for secure data analysis, is severely limited by the relatively small size of a secure enclave, a private region of the memory protected from other processes. SkSES addresses this limitation through the use of sketching data structures to support efficient secure and privacy preserving SNP analysis across individually encrypted VCF files from multiple institutions. In particular SkSES provides the users the ability to query for the "k" most significant SNPs among any set of user specified SNPs and any value of "k" - even when the total number of SNPs to be maintained is far beyond the memory capacity of the secure enclave. Results: We tested SkSES on the complete iDASH-2017 competition data set comprised of 1000 case and 1000 control samples related to an unknown phenotype. SkSES was able to identify the top SNPs with respect to the chi-squared statistic, among any user specified subset of SNPs across this data set of 2000 individually encrypted complete human genomes quickly and accurately - demonstrating the feasibility of secure and privacy preserving computation for genomic medicine via Intel's SGX. Availability: https://github.com/ndokmai/sgx-genome-variants-search Contact: cenksahi@indiana.edu


  • A Two-Level Scheme for Quality Score Compression

    by Jan Voges, Ali Fotouhi, Jörn Ostermann, Muhammed Oğuzhan Külekci

    Previous studies on quality score compression can be classified into two main lines: lossy schemes and lossless schemes. Lossy schemes enable a better management of computational resources. Thus, in practice, and for preliminary analyses, bioinformaticians may prefer to work with a lossy quality score representation. However, the original quality scores might be required for a deeper analysis of the data. Hence, it might be necessary to keep them; in addition to lossy compression this requires lossless compression as well. We developed a space-efficient hierarchical representation of quality scores, QScomp, which allows the users to work with lossy quality scores in routine analysis, without sacrificing the capability of reaching the original quality scores when further investigations are required. Each quality score is represented by a tuple through a novel decomposition. The first and second dimensions of these tuples are separately compressed such that the first-level compression is a lossy scheme. The compressed information of the second dimension allows the users to extract the original quality scores. Experiments on real data reveal that the downstream analysis with the lossy part—spending only 0.49 bits per quality score on average—shows a competitive performance, and that the total space usage with the inclusion of the compressed second dimension is comparable to the performance of competing lossless schemes.