Biology subjects

Shahar, N.

Publications and source records attributed to Shahar, N..

4 recordsLinked to original sources

bioRxiv2026

ConvergeCELL: An end-to-end platform from patient transcriptomics to therapeutic hypotheses

Translating transcriptomic data into therapeutic hypotheses remains fragmented and labor-intensive. Here we present ConvergeCELL, a platform combining a patient representation model trained on over 20 million cells across 4,479 patients, an interpretability framework for gene discovery, and a large language model-driven workflow that classifies candidates along an evidence hierarchy and constructs mechanism-of-action hypotheses. Validated on held-out cohorts spanning lupus, multiple myeloma, and sepsis across single-cell and bulk modalities, ConvergeCELL recovers known disease-associated genes at or above differential expression, machine-learning, and patient-level foundation model (PaSCient) baselines. The advantage is most pronounced for clinically validated, disease-specific drug targets: ConvergeCELL ranks TNFSF13B (Belimumab; lupus), TNFRSF17/BCMA (Belantamab; myeloma), and CXCR4 (Plerixafor; myeloma) within the top 0.3% of its gene rankings - significantly outcompeting alternative approaches. ConvergeCELL delivers an end-to-end translational workflow with state-of-the-art performance on both disease-associated gene recovery and patient-level disease classification. The pretrained ConvergeCELL patient representation model and bulk distillation module are publicly available on Hugging Face (huggingface.co/ConvergeBio/virtual-cell-patient) under the Apache 2.0 license.

bioinformatics↗

bioRxiv2024

Parasite contingency loci and the evolution of host specificity: Simple sequence repeats mediate Bartonella adaptation to a wild rodent host

Parasites can adapt to better exploit their hosts on many scales, ranging from within an infection of a single individual to series of infections spanning multiple host species. However, little is known about how the genomes of parasites in natural communities evolve when they face diverse hosts. We investigated how Bartonella bacteria that circulate in rodent communities in the dunes of the Negev Desert in Israel adapt to different species of rodent hosts. We propagated 15 Bartonella populations through infections of either a single host species (Gerbillus andersoni or Gerbillus pyramidum) or alternating between the two. After 20 rodent passages, strains with de novo mutations replaced the ancestor in most populations. Mutations in two mononucleotide simple sequence repeats (SSRs) that caused frameshifts in the same adhesin gene dominated the evolutionary dynamics. They appeared exclusively in populations that encountered G. andersoni and altered the dynamics of infections of this host. Similar SSRs in other genes are conserved and exhibit ON/OFF variation in Bartonella isolates from the Negev Desert dunes. Our results suggest that SSR-based contingency loci could be important not only for rapidly and reversibly generating antigenic variation to escape immune responses but that they may also mediate the evolution of host specificity.

evolutionary biology↗

bioRxiv2023

Harnessing the flexibility of neural networks to predict dynamic theoretical parameters underlying human choice behavior

Reinforcement learning (RL) models are used extensively to study human behavior. These rely on normative models of behavior and stress interpretability over predictive capabilities. More recently, neural network models have emerged as a descriptive modeling paradigm that is capable of high predictive power yet with limited interpretability. Here, we seek to augment the expressiveness of theoretical RL models with the high flexibility and predictive power of neural networks. We introduce a novel framework, which we term theoretical-RNN (t-RNN), whereby a recurrent neural network is trained to predict trial-by-trial behavior and to infer theoretical RL parameters using artificial data of RL agents performing a two-armed bandit task. In three studies, we then examined the use of our approach to dynamically predict unseen behavior along with time-varying theoretical RL parameters. We first validate our approach using synthetic data with known RL parameters. Next, as a proof-of-concept, we applied our framework to two independent datasets of humans performing the same task. In the first dataset, we describe differences in theoretical RL parameters dynamic among clinical psychiatric vs. healthy controls. In the second dataset, we show that the exploration strategies of humans varied dynamically in response to task phase and difficulty. For all analyses, we found better performance in the prediction of actions for t-RNN compared to the stationary maximum-likelihood RL method. We discuss the use of neural networks to facilitate the estimation of latent RL parameters underlying choice behavior. Author summaryCurrently, neural network models fitted directly to behavioral human data are thought to dramatically outperform theoretical computational models in terms of predictive accuracy. However, these networks do not provide a clear theoretical interpretation of the mechanisms underlying the observed behavior. Generating plausible theoretical explanations for observed human data is a major goal in computational neuroscience. Here, we provide a proof-of-concept for a novel method where a recurrent neural network (RNN) is trained on artificial data generated from a known theoretical model to predict both trial-by-trial actions and theoretical parameters. We then freeze the RNN weights and use it to predict both actions and theoretical parameters of empirical data. We first validate our approach using synthetic data where the theoretical parameters are known. We then show, using two empirical datasets, that our approach allows dynamic estimation of latent parameters while providing better action predictions compared to theoretical models fitted with a maximum-likelihood approach. This proof-of-concept suggests that neural networks can be trained to predict meaningful time-varying theoretical parameters.

animal behavior and cognition↗

bioRxiv2018

Solving the riddle of the evolution of Shine-Dalgarno based translation in chloroplasts

The chloroplast, a photosynthetic organelle found in all plant and algae species, originates from an ancient event in which a cyanobacterium was engulfed by a larger eukaryote. Thus, modern chloroplasts still harbor a bacterial-like genome and carry out all stages of gene expression, including mRNA translation by a 70S ribosome. However, the Shine-Dalgarno model, which predominantly regulates translation initiation by base-pairing between the ribosomal RNA and the mRNA in model bacteria genera, was reported to have ambiguous effects on chloroplast gene expression. Here we show that while the Shine-Dalgarno motif is clearly conserved in proteobacterial mRNAs, its general absence from chloroplast mRNAs is observed in cyanobacteria as well, promoting the idea that the evolutionary process of reducing the centrality of the Shine-Dalgarno mechanism began well before plastid endosymbiosis. As plastid ribosomal RNA anti-Shine-Dalgarno elements are highly similar to their bacterial counterparts, these sites alone cannot explain the decline in plastid Shine-Dalgarno generality. However, by computational simulation we show that upstream point mutations modulate the local structure of ribosomal RNA in chloroplasts, creating significantly tighter structures around the anti-Shine-Dalgarno locus, which in-turn reduce the probability of ribosome binding via the Shine-Dalgarno mechanism. To validate our model, we expressed a mCherry gene harboring a Shine-Dalgarno motif in the Chlamydomonas reinhardtii chloroplast. We show that co-expressing it with a 16S ribosomal RNA, modified according to our model, significantly enhances its expression compared to co-expression with an endogenous 16S gene.\n\nSignificance statementChloroplasts are fascinating intracellular organelles which have evolved from an ancient cyanobacterium engulfed by a larger eukaryote. Surprisingly, the canonical mechanism regulating bacterial translation initiation - Shine-Dalgarno - has been shown to play a reduced role in chloroplasts. Here, we show that mutations upstream from the anti-Shine-Dalgarno element decrease the probability of spontaneous ribosome binding by modulating the secondary structure of the ribosomal RNA. These mutations constitute a regulatory step which acclimates the Shine-Dalgarno mechanism to the translational regulation regime of chloroplasts. Interestingly, we show that these chloroplast features occur in modern cyanobacteria as well, promoting the idea that they have evolved prior to endosymbiosis.

plant biology↗