Prof. Yaron Orenstein
Computational biology, computational genomics, bioinformatics, machine learning for biology, deep learning, sequence algorithms, k-mer analysis, minimizers, universal hitting sets, sequence sketching, genome assembly, de Bruijn graphs, CRISPR/Cas9 prediction, on-target and off-target prediction, gene regulation, epigenomics, protein–DNA interactions, protein–RNA interactions, sequence-to-function modeling, high-throughput sequencing, algorithm design, graph algorithms, scalable computation, and open-source bioinformatics tools.
CV
Research
The Orenstein Computational Biology lab
The Orenstein Lab develops algorithms, computational models, and artificial intelligence (AI) tools to understand complex biological data, with a particular focus on DNA and RNA sequences and their interactions with proteins. Combining computational biology, computational genomics, bioinformatics, and machine learning for biology, the lab investigates how biological sequences determine molecular interactions and functions.
A central research theme is the prediction of gene-regulatory mechanisms, including transcription factor binding and RNA-binding protein interactions. By integrating machine learning, deep learning, neural networks, and high-throughput experimental datasets such as PBM, HT-SELEX, ChIP, and RNAcompete, the lab develops models for predicting protein–DNA and protein–RNA interactions. Its research also includes computational methods for predicting nucleic acid structures and properties, including G-quadruplexes, as well as models for assessing the functional effects of protein variants.
Another research direction focuses on CRISPR/Cas9, particularly the prediction of genome-editing efficiency across different cellular contexts. Alongside these biological applications, the lab develops efficient, scalable algorithms for genome-scale sequence analysis, including k-mer analysis, minimizers, universal hitting sets, sequence indexing, and de Bruijn graphs.
The lab’s distinctive strength lies in the close integration of theoretical computer science, algorithm design, and experimental biological data. By combining mathematically grounded and computationally efficient methods with predictive models that can be evaluated and applied to real biological systems, the lab advances our ability to analyze biological sequences, understand molecular interactions, and predict biological function.
Courses
- 80-541 – Deep Learning in Computational Biology
- 80-959 – Seminar in Computational Biology
- 80-409 – Biotechnology Seminar
Publications
Selected Publications
1. Yang, L., Orenstein, Y., Jolma, A., Yin, Y., Taipale, J., Shamir, R., & Rohs, R. (2017). Transcription factor family-specific DNA shape readout revealed by quantitative specificity models. Molecular Systems Biology, 13, 910.
https://doi.org/10.15252/msb.20167238
Demonstrates that DNA shape contributes to transcription factor binding specificity across diverse protein families, revealing how machine-learning models can uncover mechanisms of protein–DNA recognition without requiring experimentally determined atomic structures.
2. Orenstein, Y., & Shamir, R. (2014). A comparative analysis of transcription factor binding models learned from PBM, HT-SELEX and ChIP data. Nucleic Acids Research, 42(8), e63.
https://doi.org/10.1093/nar/gku117
Systematically compares transcription factor binding models derived from PBM and HT-SELEX experiments, revealing differences in predictive performance and experimental biases that are important for accurately modeling gene regulation.
3. Ben-Bassat, I., Chor, B., & Orenstein, Y. (2018). A deep neural network approach for learning intrinsic protein-RNA binding preferences. Bioinformatics, 34(17), i638–i646.
https://doi.org/10.1093/bioinformatics/bty600
Introduces DLPRB, a deep-learning approach that combines RNA sequence and structural information to predict protein–RNA interactions, substantially improving binding prediction from experimental data and providing insight into RNA-binding specificity.
4. Orenstein, Y., Wang, Y., & Berger, B. (2016). RCK: accurate and efficient inference of sequence- and structure-based protein–RNA binding models from RNAcompete data. Bioinformatics, 32(12), i351–i359.
https://doi.org/10.1093/bioinformatics/btw259
Develops RCK, a scalable k-mer-based algorithm that learns both sequence and structural preferences of RNA-binding proteins, improving the accuracy and computational efficiency of protein–RNA binding prediction from large-scale RNAcompete data.
5. Orenstein, Y., Pellow, D., Marçais, G., Shamir, R., & Kingsford, C. (2017). Designing small universal k-mer hitting sets for improved analysis of high-throughput sequencing. PLOS Computational Biology, 13(10), e1005777.
https://doi.org/10.1371/journal.pcbi.1005777
Introduces an algorithm for constructing compact universal k-mer hitting sets that outperform conventional minimizers in representing genomic sequences, enabling more efficient large-scale sequence analysis with reduced computational requirements.
Media
The Israeli Model That Could Accelerate the Development of New Antiviral Drugs | ynet, Anna Rivkin, Davidson Institute | An article about research involving Dr. Yaron Orenstein that combines machine learning and synthetic biology to predict protein–RNA interactions, with potential applications in antiviral drug development.
Last Updated Date : 15/10/2026
