Biology subjects

Yakhini, Z.

Publications and source records attributed to Yakhini, Z..

17 recordsLinked to original sources

bioRxiv2026

NinjaSeq: programmable restriction enzyme-based sequencing library preparation with random access for DNA data storage

DNA data storage allows sequences to be defined without biological constraints, yet readout workflows still depend on generic end-repair/dA-tailing chemistry. We developed NinjaSeq, a type IIS restriction endonuclease library-preparation strategy that incorporates recognition sites into primer flanks, enabling digestion to generate adapter-compatible overhangs and eliminating the need for conventional end preparation. By combining this chemistry with constrained coding that excludes internal recognition motifs, NinjaSeq produced sequencing quality and decoding performance consistent with standard protocols while reducing reagent burden and simplifying processing, including compatibility with one-pot restriction-ligation. The same sequence-directed design also enables physical random access during library preparation: targeting file-specific flanking sites enriched a desired file from a mixed pool by about sixteen-fold in a proof-of-concept experiment. These results position NinjaSeq as a practical ONT readout approach for DNA data storage. HIGHLIGHTSO_LINinjaSeq replaces end-repair/dA-tailing with REases for nanopore sequencing C_LIO_LIConstrained encoding excludes recognition motifs to protect payloads from cleavage C_LIO_LINinjaSeq achieves decoding accuracy comparable to standard library preparation C_LIO_LIDesigning file-specific RRS enables random access during library preparation C_LI

molecular biology↗

bioRxiv2024

HIPI: Spatially Resolved Multiplexed Protein Expression Inferred from H&E WSIs

Solid tumors are characterized by complex interactions between the tumor, the immune system and the microenvironment. These interactions and intra-tumor variations have both diagnostic and prognostic significance and implications. However, quantifying the underlying processes in patient samples requires expensive and complicated molecular experiments. In contrast, H&E staining is typically performed as part of the routine standard process, and is very cheap. Here we present HIPI (H&E Image Interpretation and Protein Expression Inference) for predicting cell marker expression from tumor H&E images. We process paired H&E and CyCIF images taken from serial sections of colorectal cancers to train our model. We show that our model accurately predicts the spatial distribution of several important cell markers, on both held-out tumor regions as well as new tumor samples taken from different patients. Moreover, using only the tissue image morphology, HIPI is able to colocalize the interactions between different cell types, further demonstrating its potential clinical significance.

bioinformatics↗

bioRxiv2024

Sequencing coverage analysis for combinatorialDNA-based storage systems

This study introduces a novel model for analyzing and determining the required sequencing coverage in DNA-based data storage, focusing on combinatorial DNA encoding. We explore the application of the coupon collector model for combinatorial-letter reconstruction, post-sequencing, which ensure efficient data retrieval and error reduction. We use a Markov Chain model to compute the probability of error-free reconstruction. We develop theoretical bounds on the decoding probability and use empirical simulations to validate these bounds. The work contributes to the understanding of sequencing coverage in DNA-based data storage, offering insights into decoding complexity, error correction, and sequence reconstruction. We provide a Python package that takes the code design and other message parameters as input, and then computes the required read coverage to guarantee reconstruction at a given desired confidence.

synthetic biology↗

bioRxiv2022

SPIRAL: Significant Process InfeRence ALgorithm for single cell RNA-sequencing and spatial transcriptomics

1Gene expression data is complex and may hold information regarding multiple biological processes at once. We present SPIRAL, an algorithm that uses a Gaussian statistical model to produce a comprehensive overview of a plurality of significant processes detected in single cell RNA-seq or spatial transcriptomics data. SPIRAL identifies biological processes by finding sub-matrices that consist of the subset of genes involved and the subset of cells or spots. We describe the algorithmic method, the analysis pipeline and several example results. SPIRAL is available at https://spiral.technion.ac.il/.

bioinformatics↗

bioRxiv2021

Data Storage Based on Combinatorial Synthesis of DNA Shortmers

1With the world generating digital data at an exponential rate, DNA has emerged as a promising archival medium. It offers a more efficient and long-lasting digital storage solution due to its durability, physical density, and high information capacity. Research in the field includes the development of encoding schemes, which are compatible with existing DNA synthesis and sequencing technologies. Recent studies suggest leveraging the inherent information redundancy of these technologies by using composite DNA alphabets. A major challenge in this approach involves the noisy inference process, which prevented the use of large composite alphabets. This paper introduces a novel approach for DNA-based data storage, offering a 6.5-fold increase in logical density over standard DNA-based storage systems, with near zero reconstruction error. Combinatorial DNA encoding uses a set of clearly distinguishable DNA shortmers to construct large combinatorial alphabets, where each letter represents a subset of shortmers. The nature of these combinatorial alphabets minimizes mix-up errors, while also ensuring the robustness of the system. As this paper will show, we formally define various combinatorial encoding schemes and investigate their theoretical properties, such as information density, reconstruction probabilities and required synthesis, and sequencing multiplicities. We then suggest an end-to-end design for a combinatorial DNA-based data storage system, including encoding schemes, two-dimensional error correction codes, and reconstruction algorithms. Using in silico simulations, we demonstrate our suggested approach and evaluate different combinatorial alphabets for encoding 10KB messages under different error regimes. The simulations reveal vital insights, including the relative manageability of nucleotide substitution errors over shortmer-level insertions and deletions. Sequencing coverage was found to be a key factor affecting the system performance, and the use of two-dimensional Reed-Solomon (RS) error correction has significantly improved reconstruction rates. Our experimental proof-of-concept validates the feasibility of our approach, by constructing two combinatorial sequences using Gibson assembly imitating a 4-cycle combinatorial synthesis process. We confirmed the successful reconstruction, and established the robustness of our approach for different error types. Subsampling experiments supported the important role of sampling rate and its effect on the overall performance. Our work demonstrates the potential of combinatorial shortmer encoding for DNA-based data storage, while raising theoretical research questions and technical challenges. These include the development of error correction codes for combinatorial DNA, the exploration of optimal sampling rates, and the advancement of DNA synthesis technologies that support combinatorial synthesis. Combining combinatorial principles with error-correcting strategies paves the way for efficient, error-resilient DNA-based storage solutions.

synthetic biology↗

bioRxiv2021

Assessing heterogeneity in spatial data using the HTA index with applications to spatial transcriptomics and imaging

MotivationTumour heterogeneity is being increasingly recognised as an important characteristic of cancer and as a determinant of prognosis and treatment outcome. Emerging spatial transcriptomics data hold the potential to further our understanding of tumour heterogeneity and its implications. However, existing statistical tools are not sufficiently powerful to capture heterogeneity in the complex setting of spatial molecular biology. ResultsWe provide a statistical solution, the HeTerogeneity Average index (HTA), specifically designed to handle the multivariate nature of spatial transcriptomics. We prove that HTA has an approximately normal distribution, therefore lending itself to efficient statistical assessment and inference. We first demonstrate that HTA accurately reflects the level of heterogeneity in simulated data. We then use HTA to analyse heterogeneity in two cancer spatial transcriptomics datasets: spatial RNA sequencing by 10x Genomics and spatial transcriptomics inferred from H&E. Finally, we demonstrate that HTA also applies to 3D spatial data using brain MRI. In spatial RNA sequencing we use a known combination of molecular traits to assert that HTA aligns with the expected outcome for this combination. We also show that HTA captures immune-cell infiltration at multiple resolutions. In digital pathology we show how HTA can be used in survival analysis and demonstrate that high levels of heterogeneity may be linked to poor survival. In brain MRI we show that HTA differentiates between normal ageing, Alzheimers disease and two tumours. HTA also extends beyond molecular biology and medical imaging, and can be applied to many domains, including GIS. Availability . Contactlevyalona@gmail.com zohar.yakhini@gmail.com Supplementary informationSupplementary data are available at Bioinformatics online.

bioinformatics↗

bioRxiv2020

Spatial Transcriptomics Inferred from Pathology Whole-Slide Images Using Deep Learning Links Tumor Heterogeneity to Survival in Breast and Lung Cancer

Digital analysis of pathology whole-slide images is fast becoming a game changer in cancer diagnosis and treatment. Specifically, deep learning methods have shown great potential to support pathology analysis, with recent studies identifying molecular traits that were not previously recognized on pathology H&E whole-slide images. Simultaneous to these developments, it is becoming increasingly evident that tumor heterogeneity is an important determinant of cancer prognosis and susceptibility to treatment, and should therefore play a role in the evolving practices of matching treatment protocols to patients. State of the art diagnostic procedures, however, do not provide automated methods for characterizing and/or quantifying tumor heterogeneity, certainly not in a spatial context. Further, existing methods for analyzing pathology whole-slide images from bulk measurements require many training samples and complex pipelines. Our work addresses these two challenges. First, we train deep learning models to spatially resolve bulk mRNA and miRNA expression levels on pathology whole-slide images (WSIs). Our models reach up to 0.95 AUC on held-out test sets from two cancer cohorts using a simple training pipeline and a small number of training samples. Using the inferred gene expression levels, we further develop a method to spatially characterize tumor heterogeneity. Specifically, we produce tumor molecular cartographies and heterogeneity maps of WSIs and formulate a heterogeneity index (HTI) that quantifies the level of heterogeneity within these maps. Applying our methods to breast and lung cancer slides, we show a significant statistical link between heterogeneity and survival. Our methods potentially open a new and accessible approach to investigating tumor heterogeneity and other spatial molecular properties and their link to clinical characteristics, including treatment susceptibility and survival.

bioinformatics↗

bioRxiv2020

Sequence determinants and evolution of splicing in budding yeast and related species

RNA splicing is a key process in eukaryotic gene expression. Most Intron-containing genes are constitutively spliced, hence efficient splicing of an intron is crucial for efficient gene expression. Here we use a large synthetic oligo library of ~20,000 variants to explore how different intronic sequence features affect splicing efficiency and mRNA expression levels in S. cerevisiae. Using a combinatorial design of synthetic introns we demonstrate how non-consensus splice site sequences affect splicing efficiency in each of the three splice sites. We then show that S. cerevisiae splicing machinery tends to select alternative 3 splice sites downstream of the original site, and we suggest that this tendency created a selective pressure, leading to the avoidance of cryptic splice site motifs near introns 3 ends. We further use natural intronic sequences from other yeast species, whose splicing machineries have diverged to various extents, to show how intron architectures in the various species have been adapted to the organisms splicing machinery. We suggest that the observed tendency for cryptic splicing is a result of a loss of a specific splicing factor, U2AF1. Lastly, we show that synthetic sequences containing two introns give rise to alternative RNA isoforms in S. cerevisiae, exposing intronic features that control and facilitate alternative splicing. Our study reveals novel mechanisms by which introns are shaped in evolution to allow cells to regulate their transcriptome.

systems biology↗

bioRxiv2019

Identifying triplex binding rules in vitro leads to creation of a new synthetic regulatory tool in vivo

Nature provides a rich toolbox of dynamic nucleic acid structures that are widespread in cells and affect multiple biological processes1. Recently, non-canonical structures gained renewed scientific and biotechnological interest2,3. One particularly intriguing form of such structures are triplexes4 in which a single-stranded nucleic acid molecule interacts via Hoogsteen bonds with a DNA/RNA double helix5. Despite extensive research in vitro6-9, the underlying rules for triplex formation remain debated and evidence for triplexes in vivo is circumstantial10-12. Here, we demonstrate the development of a deep-sequencing platform termed Triplex-Seq to systematically refine the DNA triplex code and identify high affinity triplex forming oligo (TFO) variants. We identified a preference for short G-rich motifs using an oligo-library with a mix of all four bases. These high-information content motifs formed specific high-affinity triplexes in a pH-independent manner and stability was increased with G-rich double-stranded molecules. We then conjugated one high-affinity and one low-affinity variant to a VP48 peptide and studied these synthetic biomolecules in mammalian cells. Using these peptide-oligo constructs (POCs), we demonstrated possible triplex-induced down-regulation activity in 544 differentially expressed genes. Our results show that deep-sequencing platforms can substantially expand our understanding of triplex binding rules, which in turn has led to the development of a functional non-genetically encoded regulatory tool for in vivo applications.

synthetic biology↗

bioRxiv2019

Overcoming the design, build, test (DBT) bottleneck for synthesis of nonrepetitive protein-RNA binding cassettes for RNA applications

The design-build-test (DBT) cycle in synthetic biology is considered to be a major bottleneck for progress in the field. The emergence of high-throughput experimental techniques, such as oligo libraries (OLs), combined with machine learning (ML) algorithms, provide the ingredients for a potential "big-data" solution that can generate a sufficient predictive capability to overcome the DBT bottleneck. In this work, we apply the OL-ML approach to the design of RNA cassettes used in gene editing and RNA tracking systems. RNA cassettes are typically made of repetitive hairpins, therefore hindering their retention, synthesis, and functionality. Here, we carried out a high-throughput OL-based experiment to generate thousands of new binding sites for the phage coat proteins of bacteriophages MS2 (MCP), PP7 (PCP), and Q{beta} (QCP). We then applied a neural network to vastly expand this space of binding sites to millions of additional predicted sites, which allowed us to identify the structural and sequence features that are critical for the binding of each RBP. To verify our approach, we designed new non-repetitive binding site cassettes and tested their functionality in U2OS mammalian cells. We found that all our cassettes exhibited multiple trackable puncta. Additionally, we designed and verified two additional cassettes, the first containing sites that can bind both PCP and QCP, and the second with sites that can bind either MCP or QCP, allowing for an additional orthogonal channel. Consequently, we provide the scientific community with a novel resource for rapidly creating functional non-repetitive binding site cassettes using one or more of three phage coat proteins with a variety of binding affinities for any application spanning bacteria to mammalian cells.

synthetic biology↗

bioRxiv2019

SOLQC : Synthetic Oligo Library Quality Control Tool.

MotivationRecent years have seen a growing number and a broadening scope of studies using synthetic oligo libraries for a range of applications in synthetic biology. As experiments are growing by numbers and complexity, analysis tools can facilitate quality control and help in assessment and inference. ResultsWe present a novel analysis tool, called SOLQC, which enables fast and comprehensive analysis of synthetic oligo libraries, based on NGS analysis performed by the user. SOLQC provides statistical information such as the distribution of variant representation, different error rates and their dependence on sequence or library properties. SOLQC produces graphical descriptions of the analysis results. The results are reported in a flexible report format. We demonstrate SOLQC by analyzing literature libraries. We also discuss the potential benefits and relevance of the different components of the analysis. Availabilityhttps://app.gitbook.com/@yoav-orlev/s/solqc/

bioinformatics↗

bioRxiv2018

Predicting Methylation from Sequence and Gene Expression Using Deep Learning with Attention

DNA methylation has been extensively linked to alterations in gene expression, playing a key role in the manifestation of multiple diseases, most notably cancer. For this reason, researchers have long been measuring DNA methylation in living organisms. The relationship between methylation and expression, and between methylation in different genomic regions is of great theoretical interest from a molecular biology perspective. Therefore, several models have been suggested to support the prediction of methylation status in samples. These models, however, have two main limitations: (a) they heavily rely on partially measured methylation levels as input, somewhat defeating the object as one is required to collect measurements from the sample of interest before applying the model; and (b) they are largely based on human mediated feature engineering, thus preventing the model from unveiling its own representations. To address these limitations we used deep learning, with an attention mechanism, to produce a general model that predicts DNA methylation for a given sample in any CpG position based solely on the sample's gene expression profile and the sequence surrounding the CpG.\n\nWe show that our model is capable of generalizing to a completely separate test set of CpG positions and subjects. Depending on gene-CpG proximity conditions, our model can attain a Spearman correlation of up to 0.8 and MAE of 0.14 for thousands of CpG sites in the test data. We also identify and analyze several motifs and genes that our model suggests may be linked to methylation activity, such as Nodal and Hand1. Moreover, our approach, and most notably the use of attention mechanisms, offers a novel framework with which to extract valuable insights from gene expression data when combined with sequence information.\n\nThe code and trained models are available at: https://github.com/YakhiniGroup/Methylation

bioinformatics↗

bioRxiv2018

Improved DNA based storage capacity and fidelity using composite DNA letters

DNA, with its remarkable density and long-term stability, is an appealing potential next generation data storage medium, most notably for long-term archiving. Megabyte scale DNA based storage was first reported in 2012. The Shannon information capacity of DNA was recently demonstrated, using fountain codes, to be [~]1.57 bit per synthesized position. However, synthesis and sequencing technologies process multiple nominally identical molecules in parallel, leading to significant information redundancies. We introduce composite DNA alphabets, using mixed DNA base types, to leverage this redundancy, enabling higher density. We develop encoding and decoding for composite DNA based storage, including error correction. Using current DNA synthesis technologies, we code 6.4 Megabyte data into composite DNA, achieving [~]25% increase in capacity as compared to literature. We further demonstrate, on smaller scales, how flexible synthesis leads to 2.7 fold increased capacity per synthesized position. Composite DNA can thus reduce costs for DNA based storage and can also serve in other applications.

synthetic biology↗

bioRxiv2018

Haplotypes associated to gene expression in breast cancer: can they lead us to the susceptibility markers?

We have undertaken a systematic haplotype analysis of the positional type of biclusters analysing samples collected from 164 breast cancer patients and 86 women with no known history of breast cancer. We present here the haplotypes and LD patterns in more than 80 genes distributed across all chromosomes and how they differ between cases and controls. We aim by this to 1) identify genes with different haplotype distribution or LD patterns between breast cancer patients and controls and 2) to evaluate the intratumoral mRNA expression patterns in breast cancer associated particularly to the cancer susceptibility haplotypes. A significant difference in haplotype distribution between cases and controls was observed for a total of 35 genes including ABCC1, AKT2, NFKB1, TGFBR2 and XRCC4. In addition we see a negative correlation between LD patterns in cases and controls for neighboring markers in 8 genes such as CDKN1A, EPHX1 and XRCC1.

genomics↗

bioRxiv2017

Synthetic protein-sensing riboswitches

We study translational regulation by a 5 UTR sequence encoding the binding site of an RNA-binding protein (RBP) in bacteria, using a reporter assay and Selective 2-hydroxyl acylation analysed by primer extension sequencing (SHAPE-Seq). We tested constructs containing a single hairpin, based on the binding sites of the coat RBPs of bacteriophages GA, MS2, PP7, and Q{beta}, positioned in the 5 UTR of a reporter gene. With specifically-bound RBP present, either weak repression or up-regulation is observed, depending on the binding site and its flanking sequence. SHAPE-Seq data for a representative construct exhibiting up-regulation, indicates a partially-folded hairpin and non-reactive upstream and downstream flanking region, which we attribute to intermediate structures that apparently blocks translation. RBP binding stabilizes the fully-folded hairpin state and thus facilitates translation, suggesting that the up-regulating constructs are RBP-sensing riboswitches. This finding is further supported by lengthening the binding-site stem, which in turn destabilizes the translationally-inactive state, and abolishes the up-regulating behavior. Finally, we found that the combination of two binding sites, positioned in the 5 UTR and gene-header of the same transcript, can yield a cooperative regulatory response. Together, we show that the interaction of an RBP with its RNA target facilitates structural changes in the RNA, which is reflected by a controllable range of binding affinities and dose response behaviors. Thus, demonstrating that RNA-RBP interactions can provide a platform for constructing gene regulatory networks that are based on translational, rather than transcriptional, regulation.

synthetic biology↗

bioRxiv2017

An in vivo binding assay for RNA-binding proteins based on repression of a reporter gene

We employ a reporter assay and Selective 2'-hydroxyl acylation analysed by primer extension sequencing (SHAPE-seq) to study translational regulation by RNA-binding proteins, in bacteria. We designed 82 constructs, each with a single hairpin based on the binding sites of the RNA-binding coat proteins of phages MS2, PP7, GA, and Q{beta}, at various positions within the N-terminus of a reporter gene. In the absence of RNA-binding proteins, the translation level depends on hairpin location, and exhibits a three-nucleotide periodicity. For hairpin positions within the initiation region, we observe strong translational repression in the presence of its cognate RNA-binding protein. In vivo SHAPE-seq results for a representative construct indicate that the repression phenomenon correlates with a wide-swath of protection, including the hairpin and extending past the ribosome binding site. Consequently, our data suggest that the protection provided by the RBP-hairpin complex inhibits ribosomal initiation. Finally, utilizing the repression phenomenon for quantifying protein-RNA binding affinity in vivo, we both observe partially contrasting results to previous in vitro and in situ studies, and additionally, show that this method can be used in a high-throughput assay for a quantitative study of protein-RNA binding in vivo.

synthetic biology↗

bioRxiv2016

Short CT-rich motifs can trigger context-specific silencing of gene expression in bacteria

We use an oligonucleotide library of over 10000 variants together with a synthetic biology approach to identify an insulation mechanism encoded within a subset of {sigma}54 promoters. Insulation manifests itself as dramatically reduced protein expression for a downstream gene that may be expressed by transcriptional read-through. The insulation we observe is strongly associated with the presence of short CT-rich motifs (3-5 bp), positioned within 25 bp upstream of the Shine-Dalgarno (SD) motif of the silenced gene. We hypothesize that insulation is effected by binding of the RBS to the upstream CT-rich motif. We provide evidence to support this hypothesis using mutations to the CT-rich motif and gene expression measurements on multiple sequence variants. Modelling is also consistent with this hypothesis. We show that the strength of the silencing, effected by insulation, depends on the location and number of CT-rich motifs encoded within the promoters. Finally, we show that in E.coli these insulator sequences are preferentially encoded within {sigma}54 promoters as compared to other promoter types, suggesting a regulatory role for these sequences in natural contexts. Our findings suggest that context-related regulatory effects may often be due to sequence-specific interactions encoded sparsely by short motifs that are not easily detected by lower throughput studies. Such short sequence-specific phenomena can be uncovered with a focused OL design that filters out the sequence noise, as exemplified herein.

systems biology↗