Biology subjects

Zuk, O.

Publications and source records attributed to Zuk, O..

10 recordsLinked to original sources

bioRxiv2026

Mutational Robustness Predicts Protein Dynamics Across Natural and Designed Proteins

Mutationally sensitive residues, where substitutions cause large stability changes, should also be dynamically rigid, because both properties arise from tight local packing and extensive contact networks. We test this prediction by defining a per-residue mutational robustness index, the standard deviation of predicted {Delta}{Delta}G values across all 19 single-amino-acid substitutions (which we compute using the structure-conditioned predictor ThermoMPNN), and correlating it with molecular-dynamics RMSF, crystallographic B-factors, and NMR-derived order parameters across [~]2,000 natural proteins, [~]400 de novo designs, and 759 NMR-characterized proteins. Robustness predicts dynamics well (median within-protein |{rho}| {approx} 0.6), approaching the AlphaFold2 pLDDT confidence score while providing complementary information: robustness explains additional variance beyond pLDDT on every dataset, with the largest gains on designed proteins. The correlation is equally strong on de novo designs that lack any evolutionary history, pointing to a biophysical effect rather than a proxy for sequence conservation. A multiple regression model using a full 20-dimensional {Delta}{Delta}G profile per residue further outperforms all scalar summaries, showing that the identity of the substituted amino acid encodes dynamical information not captured by single-number predictors. Case studies on individual proteins, including the Zika virus capsid where pLDDT fails almost entirely, show that robustness maps physically meaningful dynamical patterns onto three-dimensional structures. Mutational robustness is thus a physically interpretable probe of the local fitness landscape that complements structural confidence scores for predicting protein flexibility.

biophysics↗

bioRxiv2024

Natural killer cell function is regulated by TGF-β signaling in pregnancy and tumor progression

The immune compartment of the maternal-fetal interface must balance between supporting maternal-fetal interactions and maintaining maternal tolerance. Despite recent advances, the cellular and molecular regulators that drive maternal immune cell remodeling remain largely unknown. Using index and transcriptional single-cell sorting, we comprehensively characterized the immune compartment dynamics in the maternal-fetal interface of both human and mouse and charted the markers and functional pathways associated with these cells. Comparing immune signatures in decidua and tumors of human and mice, we identify conserved gene modules that are activated in natural killer (NK) cells of both compartments, including TFG-{beta} signaling. Genetic ablation and antibody blockade of the TGF-{beta} pathway in NK cells resulted in enhanced anti-tumor immunity.

immunology↗

bioRxiv2023

Selecting Covariates for Genome-Wide Association Studies

The choice of which covariates to include in a Genome-Wide Association Study (GWAS) is important since it affects the ability to detect true association signal of variants, to correct for confounders and avoid false positives, and the running time of the analysis. Commonly used covariates include age, sex, genotyping batches, genotyping array type, as well as an arbitrary number of Principal Components (PCs) used to adjust for population structure. Despite the importance of this issue, there is no consensus or clear guidelines for the right choice of covariates. Therefore, studies typically employ heuristics for their choice with no clear justification. Here, we explore the dependence of the GWAS analysis results on the choice of covariates for a wide range of quantitative and binary human phenotypes. We propose guidelines for covariates choice based on the phenotypes type (quantitative vs. disease), the heritability, and the disease prevalence, with the goal of maximizing the statistical power to detect true associations and fit accurate polygenic scores while avoiding spurious associations and minimizing computation time. We analyze 36 traits in the UK-Biobank dataset. We show that the genotype batch and assessment center can be safely removed as covariates, thus significantly reducing the GWAS computational burden for these traits.

genetics↗

bioRxiv2022

Selecting Chromosomes for Polygenic Traits

We define and study the problem of genomic block selection for multiple complex traits. In this problem, one constructs a genome by selecting different genomic parts (e.g. chromosomes) from different source genomes. The constructed genome is associated with a vector of polygenic scores, obtained by summing the polygenic scores of the different genomic parts, and the goal is to minimize a given loss function of this vector. The problem is motivated by several emerging technologies: chromosome substitution lines in crop breeding, where chromosomal segments from wild relatives are combined to improve polygenic traits such as yield and stress tolerance; chromosome transfer between yeast strains for optimizing complex industrial phenotypes; and chromosomal transplantation technologies in mammalian cells. We suggest and study several natural loss functions relevant for both quantitative and threshold traits, and show that the problem is NP-complete even for a single trait and two copies, yet only weakly so, being pseudo-polynomially solvable for any fixed number of traits. We propose three algorithms with complementary roles: a Branch-and-Bound algorithm that returns the certified global optimum for any monotone loss, a fast Block-Coordinate-Descent (BCD) heuristic with random restarts that applies to any loss, and a semidefinite-programming (SDP) relaxation that provides a certified lower bound on the optimal loss for quadratic losses, and hence an optimality-gap bound when paired with the BCD solution, empirically tight in our experiments. Using the infinitesimal model for genetic architecture, we further derive, for linear losses, a closed-form approximation for the expected gain of block selection relative to random selection across multiple traits. On yeast-scale simulations BCD matches the certified Branch-and-Bound optimum on 100% of threshold-loss instances at 466x the speed, attains a certified optimality gap of at most {approx}10% of the SDP lower bound for stabilizing-loss instances, and the realized gain roughly matches the analytic prediction.

genetics↗

bioRxiv2020

Utility of polygenic embryo screening for disease depends on the selection strategy

Polygenic risk scores (PRSs) have been offered since 2019 to screen in vitro fertilization embryos for genetic liability to adult diseases, despite a lack of comprehensive modeling of expected outcomes. Here we predict, based on the liability threshold model, the expected reduction in complex disease risk following polygenic embryo screening for a single disease. Our main finding is that a strong determinant of the potential utility of such screening is the selection strategy, a factor that has not been previously studied. Specifically, when only embryos with a very high PRS are excluded, the achieved risk reduction is minimal. In contrast, selecting the embryo with the lowest PRS can lead to substantial relative risk reductions, given a sufficient number of viable embryos. For example, a relative risk reduction of {approx}50% for schizophrenia can be achieved by selecting the embryo with the lowest PRS out of five viable embryos. We systematically examine the impact of several factors on the utility of screening, including the variance explained by the PRS, the number of embryos, the disease prevalence, the parental PRSs, and the parental disease status. When quantifying the utility, we consider both relative and absolute risk reductions, as well as population-averaged and per-couple risk reductions. We also examine the risk of pleiotropic effects. Finally, we confirm our theoretical predictions by simulating "virtual" couples and offspring based on real genomes from schizophrenia and Crohns disease case-control studies. We discuss the assumptions and limitations of our model, as well as the potential emerging ethical concerns.

genetics↗

bioRxiv2020

Deep Homology-Based Protein Contact-Map Prediction

Prediction of Proteins three dimensional structure and their contact maps from their amino-acid sequences is a fundamental problem in structural computational biology. The structure and contacts shed light on protein function, enhance our basic understanding of their molecular biology and may potentially aid in drug design. In recent years we have seen significant progress in protein contact map prediction from Multiple Sequence Alignments (MSA) of the target protein and its homologous, using signals of co-evolution and applying deep learning methods. Homology modelling is a popular and successful approach, where the structure of a protein is determined using information from known template structures of similar proteins, and has been shown to improve prediction even in cases of low sequence identity. Motivated by these observations, we developed Periscope, a method for homology-assisted contact map prediction using a deep convolutional network. Our method automatically integrates the co-evolutionary information from the MSA, and the physical contact information from the template structures. We apply our method to families of CAMEO and membrane proteins, and show improved prediction accuracy compared to the MSA-only based method RaptorX. Finally, we use our method to improve the subsequent task of predicting the proteins three dimensional structure based on the (improved) predicted contact map, and show initial promising results in this task too - our overall accuracy is comparable to the template-based Modeller software, yet the two methods are complementary and succeed on different targets.

bioinformatics↗

bioRxiv2019

Screening human embryos for polygenic traits has limited utility

Genome-wide association studies have led to the development of polygenic score (PS) predictors that explain increasing proportions of the variance in human complex traits. In parallel, progress in preimplantation genetic testing now allows genome-wide genotyping of embryos generated via in vitro fertilization (IVF). Jointly, these developments suggest the possibility of screening embryos for polygenic traits such as height or cognitive function. There are clear ethical, legal, and societal concerns regarding such a procedure, but these cannot be properly discussed in the absence of data on the expected outcomes of screening. Here, we use theory, simulations, and real data to evaluate the potential gain of PS-based embryo selection, defined as the expected difference in trait value between the top-scoring embryo and an average, unselected embryo. We observe that the gain increases very slowly with the number of embryos, but more rapidly with increased variance explained by the PS. Given currently available polygenic predictors and typical IVF yields, the average gain due to selection would be {approx}2.5cm if selecting for height, and {approx}2.5 IQ (intelligence quotient) points if selecting for cognitive function. These mean values are accompanied by wide confidence intervals; in real data drawn from nuclear families with up to 20 offspring each, we observe that the offspring with the highest PS for height was the tallest only in 25% of the families. We discuss prospects and limitations of PS-based embryo selection for the foreseeable future.

genetics↗

bioRxiv2018

Substantial Batch Effects in TCGA Exome Sequences Undermine Pan-Cancer Analysis of Germline Variants

BackgroundIn recent years, research on cancer predisposition germline variants has emerged as a prominent field. The identity of somatic mutations is based on a reliable mapping of the patient germline variants. In addition, the statistics of germline variants frequencies in healthy individuals and cancer patients is the basis for seeking candidates for cancer predisposition genes. The Cancer Genome Atlas (TCGA) is one of the main sources of such data, providing a diverse collection of molecular data including deep sequencing for more than 30 types of cancer from >10,000 patients.\n\nMethodsOur hypothesis in this study is that whole exome sequences from healthy blood samples of cancer patients are not expected to show systematic differences among cancer types. To test this hypothesis, we analyzed common and rare germline variants across six cancer types, covering 2,241 samples from TCGA. In our analysis we accounted for inherent variables in the data including the different variant calling protocols, sequencing platforms, and ethnicity.\n\nResultsWe report on substantial batch effects in germline variants associated with cancer types. We attribute the effect to the specific sequencing centers that produced the data. Specifically, we measured 30% variability in the number of reported germline variants per sample across sequencing centers. The batch effect is further expressed in nucleotide composition and variant frequencies. Importantly, the batch effect causes substantial differences in germline variant distribution patterns across numerous genes, including prominent cancer predisposition genes such as BRCA1, RET, MAX, and KRAS. For most of known cancer predisposition genes, we found a distinct batch-dependent difference in germline variants.\n\nConclusionTCGA germline data is exposed to strong batch effects with substantial variabilities among TCGA sequencing centers. We claim that those batch effects are consequential for numerous TCGA pan-cancer studies. In particular, these effects may compromise the reliability and the potency to detect new cancer predisposition genes. Furthermore, interpretation of pan-cancer analyses should be revisited in view of the source of the genomic data after accounting for the reported batch effects.

cancer biology↗

bioRxiv2018

Conservation and divergence in modules of the transcriptional programs of the human and mouse immune systems

Studies in mouse have shed important light on human hematopoietic differentiation and disease. However, substantial differences between the two species often limit the translation of findings from mouse to human.\n\nHere, we compare previously defined modules of co-expressed genes in human and mouse immune cells based on compendia of genome-wide profiles. We show that the overall modular organization of the transcriptional program is conserved. We highlight modules of co-expressed genes in one species that dissolve or split in the other species. Many of the associated regulatory mechanisms - as reflected by computationally inferred trans regulators, or enriched cis-regulatory elements - are conserved between the species. Nevertheless, the degree of conservation in regulatory mechanism is lower than that of expression, suggesting that distinct regulation may underlie some of the conserved transcriptional responses.

bioinformatics↗

bioRxiv2018

Enhancer Identification using Transfer and Adversarial Deep Learning of DNA Sequences

Enhancer sequences regulate the expression of genes from afar by providing a binding platform for transcription factors, often in a tissue-specific or context-specific manner. Despite their importance in health and disease, our understanding of these DNA sequences, and their regulatory grammar, is limited. This impairs our ability to identify new enhancers along the genome, or to understand the effect of enhancer mutations and their role in genetic diseases.\n\nWe trained deep Convolutional Neural Networks (CNN) to identify enhancer sequences in multiple species. We used multiple biological datasets, including simulated sequences, in vivo binding data of single transcription factors and genome-wide chromatin maps of active enhancers in 17 mammalian species. Our deep networks obtained high classification accuracy by combining two training strategies: First, training on enhancers vs. non-enhancer background sequences, we identified short (1-4bp) low-complexity motifs. Second, by replacing the negative training set by adversarial k-order random shuffles of enhancer sequences (thus maintaining base composition while shuttering longer motifs, including transcription factor binding sites), we identified a set of biologically meaningful motifs, unique to enhancers. In addition, classification performance improved when combining positive data from all species together, showing a shared mammalian regulatory architecture.\n\nOur results demonstrate that design of adversarial training data, and transfer of learned parameters between networks trained on different species/datasets improve the overall performance and capture biologically meaningful information in the parameters of the learned network.\n\nContact: or.zuk@mail.huji.ac.il, tommy@cs.huji.ac.il

bioinformatics↗