SEARCH · Search bioRxiv

Results for “Systems Biology”

Search indexed bioRxiv preprints in genomics, neuroscience, cell biology and bioinformatics. Read source abstracts and check manuscript versions; preprints are not peer reviewed.

Quote a phrase for an exact phrase match. Source license links do not imply unrestricted reuse.

At least 1,639 records · Page 91Linked to original sources

bioRxiv2026

Single-nucleotide resolution profiling reveals dynamic and site-specific m⁶A regulation during human myogenesis

BackgroundMyogenic differentiation of muscle stem cells required for skeletal muscle growth and regeneration relies on the precise spatiotemporal coordination of multiple gene-expression programs. Over the past decades, considerable progress has been made in defining the transcriptional networks that govern myogenesis. In particular, myogenic transcription factors integrate extrinsic and intrinsic cues to drive skeletal muscle development and regeneration. In contrast, post-transcriptional layers of regulation have remained comparatively underexplored. Among these, the epitranscriptome - comprising more than 170 chemical RNA modifications - has recently emerged as a major regulatory axis in cellular differentiation. N-methyladenosine (mA), the most abundant internal modification of mRNA, is now recognized as a key regulator of stem-cell fate decisions across multiple biological systems. By dynamically modulating RNA stability, translation, and other processing steps, mA enables precise fine-tuning of gene-expression programs in response to physiological cues. These effects are mediated through mA-binding RNA-binding proteins (RBPs), also referred to as "readers", which selectively recognize methylated transcripts and translate the modification into functional cellular outcomes. Despite its central role in other tissue contexts, the contribution of mA to skeletal muscle physiology - and particularly to human muscle stem cell biology - remains poorly characterized. This gap represents a critical and largely unexplored frontier in muscle biology. MethodsTo define the epitranscriptome landscape during the myogenic differentiation of human skeletal muscle cells, we performed GLORI on proliferating and differentiated human primary myoblasts. This recently developed approach enables detection of mA at single-nucleotide resolution together with quantitative measurement of methylation stoichiometry at individual sites. mA-modified positions were mapped across the transcriptome in both cellular states and integrated with matched RNA-seq datasets data to contextualise mA features relative to transcript abundance. To place these findings in a broader context, we compared our human mA maps with previously published mouse muscle cell datasets and overlaid them with available CLIP-seq datasets of mA RNA-binding proteins. ResultsWe identified tens of thousands of high-confidence mA sites in proliferating and differentiated human myogenic cells. While many sites were shared between states, myogenic differentiation was accompanied by epitranscriptomic remodeling, affecting both the distribution and stoichiometry of mA marks. Transcripts exhibiting differential mA regulation were enriched for pathways related to myogenic development, cytoskeletal remodeling, signalling, nucleic-acid-associated processes and energy metabolism. Joint analysis of mA site number, methylation stoichiometry, and expression levels indicated the presence of distinct mA regulatory frameworks at the transcript level. Genes associated with myogenesis and regulatory pathways were distributed across these categories, consistent with the coexistence of multiple regulatory programs during myogenic differentiation. In addition, cross-species comparisons revealed the presence of human-specific methylation patterns. Integration with published mA RNA-binding protein datasets uncovered distinct subsets of mA sites with varying binding profiles, suggesting state-dependent interpretation of mA marks. Together, these data constitute, to our knowledge, the first single-nucleotide-resolution atlas of mA methylation in human skeletal muscle, revealing a previously uncharacterized epitranscriptomic landscape and providing a foundational resource for studies of human myogenesis.

molecular biology↗

bioRxiv2019

Pathway Commons: 2019 Update

Pathway Commons (https://www.pathwaycommons.org) is an integrated resource of publicly available information about biological pathways including biochemical reactions, assembly of biomolecular complexes, transport and catalysis events and physical interactions involving proteins, DNA, RNA, and small molecules (e.g., metabolites and drug compounds). Data is collected from multiple providers in standard formats, including the Biological Pathway Exchange (BioPAX) language and the Proteomics Standards Initiative Molecular Interactions format, and then integrated. Pathway Commons provides biologists with (1) tools to search this comprehensive resource, (2) a download site offering integrated bulk sets of pathway data (e.g., tables of interactions and gene sets), (3) reusable software libraries for working with pathway information in several programming languages (Java, R, Python, and Javascript), and (4) a web service for programmatically querying the entire dataset. Visualization of pathways is supported using the Systems Biological Graphical Notation (SBGN). Pathway Commons currently contains data from 22 databases with 4,794 detailed human biochemical processes (i.e., pathways) and [~]2.3 million interactions. To enhance the usability of this large resource for end-users, we develop and maintain interactive web applications and training materials that enable pathway exploration and advanced analysis.

bioinformatics↗

bioRxiv2026

The recoverable resolution of cellular perturbation-response prediction

Virtual cell models aim to predict how cells respond to perturbations, yet accurate prediction of post-perturbation gene expression can conceal failures to recover context-specific responses. Here we identify context geometry compression (CGC), in which models preserve cellular identity and shared perturbation effects while collapsing differences in how the same perturbation acts across contexts. Across biological systems, reproducible context-specific responses remained difficult to recover; in a large drug-response matrix, this persisted even with nearly complete perturbation coverage. The same drug-response predictions achieved higher normalized recovery in predefined pathway summaries than in gene-level response profiles, while coarse response programs were more predictable in human lymphoblastoid cell lines. Compact baseline transcriptomic representations retained most full-RNA predictive performance for patient-specific ex vivo drug responses. Together, these findings establish recoverable biological resolution, rather than measurement dimensionality, as a criterion for evaluating virtual cell predictions and matching model outputs to the biological detail supported by evidence.

bioinformatics↗

bioRxiv2024

Rigosertib reverses hypertrophic cardiomyopathy in RAF1-associated Noonan syndrome

BackgroundRASopathies constitute a group of rare genetic disorders caused by mutations in genes that reside along the canonical Ras/MAPK signaling pathway, affecting cell growth and differentiation. These syndromes, which include conditions like Noonan syndrome (NS), are characterized by developmental delays, distinctive facial dysmorphia, and a variety of cardiac defects, notably hypertrophic cardiomyopathy (HCM). Despite their prevalence and impact, therapeutic options for RASopathies remain limited. Rigosertib, a novel dual Ras/MAPK and PI3K/AKT pathway inhibitor, is currently in clinical trials for treatment of melanoma and recessive dystrophic epidermolysis bullosa. Here, we identify rigosertib as a candidate therapy for RAF1-associated HCM. Methods and ResultsOur Drosophila screen of clinically relevant drugs and compounds identified rigosertib as broadly effective across a panel of transgenic RASopathy fly transgenic models, indicating that rigosertib may be effective against multiple disease isoforms. Analysis of a Drosophila model targeting a RAF1L613V transgene to the heart found that rigosertib reduced aspects of cardiac hypertrophy. Rigosertib treatment prevented or regressed cellular hypertrophy in human induced pluripotent stem cell-(iPSC-) derived cardiomyocytes homozygous for the NS-associated RAF1S257L allele. We extended these findings to a mammalian model, using Raf1L613V/+ KI mice to explore the therapeutic implications of rigosertib on RAF1-driven HCM. Longitudinal six-week treatment with rigosertib in these mice resulted in significant improvement in left ventricular chamber dimension and posterior wall thickness, total heart mass, size of individual cardiomyocytes (CMs), as well as reversal of cardiac hypertrophy. Rigosertib treatment also led to normalized fetal gene expression and inhibition of ERK and AKT pathway activities in primary CMs isolated from Raf1L613V/+ mice. Cardiac function, as assessed by echocardiography, showed significant improvement in ejection fraction and fractional shortening, with molecular studies confirming downregulation of hypertrophic markers and signaling pathways. Together with the Drosophila data, these mammalian results support the potential and use for rigosertib to reverse pathological hypertrophy in NS through targeted pathway inhibition in patients. Moreover, in addition to its effects in the heart, rigosertib treatment in mice also significantly improved other NS-associated syndromic features, including increasing bone growth and correcting craniofacial abnormalities. ConclusionsTaken together, our findings suggest rigosertib effectively normalizes and reverses RASopathy-associated HCM as well as other NS-associated syndromic features, supporting its potential for development as a promising treatment for RAF1-associated HCM and, potentially, other RASopathies-dependent pathologies. This study not only highlights the therapeutic potential of rigosertib but also demonstrates the utility of an integrated approach using Drosophila, iPSC and mammalian models to elucidate drug effects across complex biological systems.

molecular biology↗

bioRxiv2024

scDecorr - Feature decorrelation representation learning with domain adaptation enables self-supervised alignment of multiple single-cell experiments

Single-cell RNA sequencing (scRNA-seq) has revolutionized our understanding of cellular het-erogeneity in complex biological systems. However, analyzing and integrating scRNA-seq data poses unique computational challenges due to sparsity, high variability, and technical batch effects. Here, we propose a novel framework called scDecorr for robust representation learning and data integration for scRNA-seq analysis. Our approach leverages the idea of feature decorrelation-based self-supervised learning (SSL) to obtain efficient low-dimensional representations of individual cells without relying on negative samples. By maximizing similarity among distorted embeddings while decorrelating their components, scDecorr captures the biological signature while eliminating technical noise. Furthermore, scDecorr incorporates unsupervised domain adaptation to bridge the gap between batches with different probability distributions, enabling effective integration of scRNA-seq data from diverse sources. Our framework achieves domain-invariant representations by learning cell embeddings independently across domains and employing domain-specific batch normalization. We evaluate scDecorr on a variety of single cell datasets and demonstrate its ability to integrate batches without losing the inherent biological variance, thereby facilitating optimal clustering. The representations generated by scDecorr also exhibit robustness in label transfer tasks, allowing for effective transfer of cell-type labels from reference to query datasets. Overall, scDecorr offers a powerful tool for efficient analysis and integration of large and complex scRNA-seq datasets, advancing our understanding of cellular processes and disease mechanisms. The code is available here https://github.com/hayatlab/scdecorr.

bioinformatics↗

bioRxiv2024

Creating and leveraging bespoke large-scale knowledge graphs for comparative genomics and multi-omics drug discovery with SocialGene

The rapid expansion of multi-omics data has transformed biological research, offering unprecedented opportunities to explore complex genomic relationships across diverse organisms. However, the vast volume and heterogeneity of these datasets presents significant challenges for analyses. Here we introduce SocialGene, a comprehensive software suite designed to collect, analyze, and organize multi-omics data into structured knowledge graphs, with the ability to handle small projects to repository-scale analyses. Originally developed to enhance genome mining for natural product drug discovery, SocialGene has been effective across various applications, including functional genomics, evolutionary studies, and systems biology. SocialGenes concerted Python and Nextflow libraries streamline data ingestion, manipulation, aggregation, and analysis, culminating in a custom Neo4j database. The software not only facilitates the exploration of genomic synteny but also provides a foundational knowledge graph supporting the integration of additional diverse datasets and the development of advanced search engines and analyses. This manuscript introduces some of SocialGenes capabilities through brief case studies including targeted genome mining for drug discovery, accelerated searches for similar and distantly related biosynthetic gene clusters in biobank-available organisms, integration of chemical and analytical data, and more. SocialGene is free, open-source, MIT-licensed, designed for adaptability and extension, and available from github.com/socialgene.

bioinformatics↗

bioRxiv2026

MorphCell learns reconstructable shape representations with explicit scale control for 3D cell morphology

Three-dimensional (3D) cell morphology provides a measurable phenotype of cellular state and function, but representing it across biological systems and imaging modalities remains difficult. Existing approaches do not jointly provide transferable representations, complete surface reconstruction, geometric interpretation and explicit control of physical scale. Here we introduce MorphCell, a self-supervised framework that uses cell-surface point clouds to learn shape-driven representations independently of physical scale. MorphCell combines cross-view reconstruction pretraining with spherical self-reconstruction. The former captures geometric relationships between surface regions, whereas the latter recovers complete 3D morphology from individual representations. Pretrained on non-biological object surfaces, MorphCell transfers to cellular datasets without biological task-specific training and outperforms existing point-cloud representations in morphological classification. The learned representations capture both global contour and local surface variation, enabling geometric interpretation through reconstruction, biophysical descriptors and saliency analysis. By retaining physical scale as a separate variable for controlled fusion, MorphCell further reveals that shape and scale contribute differently across biological distinctions. This framework provides a general approach for representing, reconstructing and interpreting 3D cellular morphology across imaging modalities and biological contexts.

cell biology↗

bioRxiv2022

RNA cytometry of single-cells using semi-permeable microcapsules

Analytical tools for gene expression profiling of individual cells are critical for studying complex biological systems. However, the techniques enabling rapid measurements of gene expression on thousands of single-cells are lacking. Here, we report a high-throughput RNA cytometry for digital profiling of single-cells isolated in liquid droplets enveloped by a thin semi-permeable membrane (microcapsules). Due to selective permeability of the membrane, the desirable enzymes and reagents can be loaded, or replaced, in the microcapsule at any given step by simply changing the reaction buffer in which the microcapsules are dispersed. Therefore, complex molecular biology workflows can be readily adapted to conduct nucleic acid analysis on encapsulated mammalian cells, bacteria, or other biological species. The microcapsules support sequential multi-step enzymatic reactions, and remain intact under different biochemical conditions, freezing, thawing and thermocycling. Combining microcapsules with conventional FACS provides a high-throughput approach for conducting RNA cytometry of individual cells based on their digital gene expression signature.

molecular biology↗

bioRxiv2022

Influence of pump laser fluence on ultrafast structural changes in myoglobin

High-intensity femtosecond pulses from an X-ray free-electron laser enable pump probe experiments for investigating electronic and nuclear changes during light-induced reactions. On time scales ranging from femtoseconds to milliseconds and for a variety of biological systems, time-resolved serial femtosecond crystallography (TR-SFX) has provided detailed structural data for light-induced isomerization, breakage or formation of chemical bonds and electron transfer1. However, all ultra-fast TR-SFX studies to date have employed such high pump laser energies that several photons were nominally absorbed per chromophore2-14. As multiphoton absorption may force the protein response into nonphysiological pathways, it is of great concern15 whether this experimental approach16 allows valid inferences to be drawn vis-a-vis biologically relevant single-photon-induced reactions17. Here we describe ultrafast pump-probe SFX experiments on photodissociation of carboxymyoglobin, showing that different pump laser fluences yield markedly different results. In particular, the dynamics of structural changes and observed indicators of the mechanistically important coherent oscillations of the Fe-CO bond distance (predicted by recent quantum wavepacket dynamics15) are seen to depend strongly on pump laser energy. Our results confirm both the feasibility and necessity of performing TR-SFX pump probe experiments in the linear photoexcitation regime. We consider this to be a starting point for reassessing design and interpretation of ultrafast TR-SFX pump probe experiments16 such that biologically relevant insight emerges.

biophysics↗

bioRxiv2026

Learning protein function through autonomous experimental interaction

Biological AI learns primarily from existing observations, but many questions cannot be answered from available data alone. Here we show that AI can instead acquire knowledge by acting directly on biological systems and learning from the consequences. We developed a closed-loop framework in which autonomous agents design protein variants, construct and characterize them in a robotic laboratory, learn from the resulting experimental feedback, and decide what experiments to perform next. We then allowed the system to operate continuously and without human intervention for approximately one month, during which multiple agents independently explored protein sequence space while learning from shared experimental experience. Applied to glycoside hydrolases, the agents discovered enzymes with substantially altered substrate specificity toward non-native sugars and progressively learned the structure of the underlying sequence-function landscape. The resulting experimental experience also revealed determinants of substrate specificity and protein expression that were not specified as learning objectives. These results demonstrate that AI can autonomously interact with biology over extended periods to acquire knowledge through experience, establishing a framework for biological discovery driven by continuous experimental interaction.

bioengineering↗

bioRxiv2026

HoloCell: A Generative Foundation Model for Holistic Cellular Modeling

Single-cell multi-omics technologies have recently advanced to enable the profiling of epigenomic, transcriptomic, and proteomic layers within individual cells, offering new opportunities to characterize cellular states as integrated biological systems. However, developing a unified framework that can seamlessly integrate diverse omics modalities and remain robust to heterogeneous modality missingness remains challenging. Existing methods are often designed for specific modalities or modality pairs, relying on dataset-specific training or paired measurements. Here we present HoloCell, to our knowledge the first generative foundation model for joint representation learning and generative modeling across all three major single-cell omics modalities, i.e., epigenomics, transcriptomics, and proteomics. HoloCell contains over 860 million parameters and is pretrained on the Human-Multi-Omics-Corpus, which comprises approximately 468 million single-cell profiles across these three omics layers, corresponding to over 425 billion tokens. HoloCell introduces a a simple yet biologically motivated hierarchical tokenization strategy that encodes cis-regulatory elements, genes, and proteins as structured tokens within a shared modeling framework. We evaluated HoloCell across single-omics representation learning, paired multi-omics integration, unpaired multi-omics alignment, and cross-modal generation via iterative diffusion and remasking, demonstrating its superior performance and flexibility across diverse omics tasks. From a representation perspective, HoloCell provides a unified digital mapping of cellular states across multiple omics layers, capturing cell heterogeneity as an integrated system. From a generation perspective, its iterative diffusion and remasking frame-work permits flexible generation orders beyond fixed left-to-right causality, enabling in silico simulation of multi-omics information flow. Together, these capabilities position HoloCell as a versatile foundation model toward the emerging concept of a virtual cell, offering both systematic characterization and generative simulation of cellular systems within a unified framework.

bioinformatics↗

bioRxiv2022

DeepGAMI: Deep biologically guided auxiliary learning for multimodal integration and imputation to improve phenotype prediction

Genotype-phenotype association is found in many biological systems, such as brain-related diseases and behavioral traits. Despite the recent improvement in the prediction of phenotypes from genotypes, they can be further improved and explainability of these predictions remains challenging, primarily due to complex underlying molecular and cellular mechanisms. Emerging multimodal data enables studying such mechanisms at different scales from genotype to phenotypes involving intermediate phenotypes like gene expression. However, due to the black-box nature of many machine learning techniques, it is challenging to integrate these multi-modalities and interpret the biological insights in prediction, especially when some modality is missing. Biological knowledge has recently been incorporated into machine learning modeling to help understand the reasoning behind the choices made by these models. To this end, we developed DeepGAMI, an interpretable deep learning model to improve genotype-phenotype prediction from multimodal data. DeepGAMI uses prior biological knowledge to define the neural network architecture. Notably, it embeds an auxiliary-learning layer for cross-modal imputation while training the model from multimodal data. Using this pre-trained layer, we can impute latent features of additional modalities and thus enable predicting phenotypes from a single modality only. Finally, the model uses integrated gradient to prioritize multimodal features and links for phenotypes. We applied DeepGAMI to multiple emerging multimodal datasets: (1) population-level genotype and bulk-tissue gene expression data for predicting schizophrenia, (2) population-level genotype and gene expression data for predicting clinical phenotypes in Alzheimers Disease, (3) gene expression and electrophysiological data of single neuronal cells in the mouse visual cortex, and (4) cell-type gene expression and genotype data for predicting schizophrenia. We found that DeepGAMI outperforms existing state-of-the-art methods and provides a profound understanding of gene regulatory mechanisms from genotype to phenotype, especially at cellular resolution. DeepGAMI is an open-source tool and is available at https://github.com/daifengwanglab/DeepGAMI.

bioinformatics↗

bioRxiv2022

The seeker R package: simplified fetching and processing of transcriptome data

Transcriptome data have become invaluable for interrogating biological systems. Preparing a transcriptome dataset for analysis, particularly an RNA-seq dataset, entails multiple steps and software programs, each with its own command-line interface (CLI). Although these CLIs are powerful, they often require shell scripting for automation and parallelization, which can have a high learning curve, especially when the details of the CLIs vary from one tool to another. However, many individuals working with transcriptome data are already familiar with R due to the plethora and popularity of R-based tools for analyzing biological data. Thus, we developed an R package called seeker for simplified fetching and processing of RNA-seq and microarray data. Seeker is a wrapper around various existing tools, and provides a standard interface, simple parallelization, and detailed logging. Seekers primary output--sample metadata and gene expression values based on Entrez or Ensembl Gene IDs--can be directly plugged into a differential expression analysis. To maximize reproducibility, seeker is available as a standalone R package and in a Docker image that includes all dependencies, both of which are accessible at https://seeker.hugheylab.org.

bioinformatics↗

bioRxiv2023

Selective sweep probabilities in spatially expanding populations

Evolution during range expansions shapes biological systems from microbial communities and tumours up to invasive species. A fundamental question is whether, when a beneficial mutation arises during a range expansion, it will evade clonal interference and sweep through the population to fixation. However, most theoretical investigations of range expansions have been confined to regimes in which selective sweeps are effectively impossible, while studies of selective sweeps have either assumed constant population size or have ignored spatial structure. Here we use mathematical modelling and analysis to investigate selective sweep probabilities in the alternative yet biologically relevant scenario in which mutants can outcompete and displace a slowly spreading wildtype. Assuming constant radial expansion speed, we derive probability distributions for the arrival time and location of the first surviving mutant and hence find surprisingly simple approximate and exact expressions for selective sweep probabilities in one, two and three dimensions, which are independent of mutation rate. Namely, the selective sweep probability is approximately (1 - cwt/cm)d, where cwt and cm are the wildtype and mutant radial expansion speeds, and d the spatial dimension. Using agent-based simulations, we show that our analytical results accurately predict selective sweep frequencies in the two-dimensional spatial Moran process. We further compare our results with those obtained for alternative growth laws. Parameterizing our model for human tumours, we find that selective sweeps are predicted to be rare except during very early solid tumour growth, thus providing a general, pan-cancer explanation for findings from recent sequencing studies.

evolutionary biology↗

bioRxiv2025

scMomer: A modality-aware pretraining framework for single-cell multi-omics modeling under missing modality conditions

Foundation models offer new opportunities to capture cellular behavior from large-scale single-cell data. However, their development has been greatly constrained due to the limited availability of multi-omics profiles. Consequently, most models are designed for a single modality (e.g. scRNA-seq, or scATAC-seq, etc.), restricting their ability to capture the diversity of heterogeneous biological systems. Here, we introduce scMomer, a modality-aware pretraining framework designed for multi-modal representation learning under missing modality conditions. scMomer adopts a three-stage pretraining strategy that learns unimodal cell representations, models joint representations from multi-omics data, and distills multi-modal knowledge to enable multi-omics-like representations from unimodal input. Its modality-specific architecture and three-stage pretraining strategy enable effective learning under missing modality conditions and help capture cellular heterogeneity. Through extensive experiments, scMomer generates biologically meaningful embeddings and outperforms state-of-the-art unimodal approaches across diverse gene-level and cell-level downstream tasks, including cross-modality translation, gene function prediction, cell annotation, drug response prediction, and perturbation prediction. Overall, these results demonstrate that scMomer serves as a robust, generalizable, and scalable foundation for single-cell multi-modal analysis under missing modality conditions.

bioinformatics↗

bioRxiv2026

Bringing calorimetry (back) to life

Micro-calorimetry offers significant potential as a quantitative method for studying the structure and function of biological systems, for instance, by probing the excess heat released by cellular or sub-cellular structures, isothermal or not, when external parameters change. We present the conceptual framework of nonequilibrium calorimetry, and as illustrations, we compute the heat capacity of biophysical models with few degrees of freedom related to ciliar motion (rowing model) and molecular motor motion (flashing ratchets). Our quantitative predictions reveal intriguing dependencies of the (nonequilibrium) heat capacity as a function of relevant biophysical parameters, which can even take negative values as a result of biological activity.

biophysics↗

bioRxiv2024

Genetic designs for stochastic and probabilistic biocomputing

The programming of computations in living cells can be done by manipulating information flows within genetic networks. Typically, a single bit of information is encoded by a single genes steady state expression. Expression is discretized into high and low levels that correspond to 0 and 1 logic values, analogous to the high and low voltages in electronic logic circuits. However, the processes of molecular signaling and computation in living systems challenge this computational paradigm with their dynamic, stochastic and continuous operation. Although there is a good understanding of these phenomena in genetic networks, and there are already stochastic and probabilistic models of computation which can take on these challenges, there is currently a lack of work which puts both together to implement computations tailored to these features of living matter. Here, we design genetic networks for stochastic and probabilistic computing paradigms and develop the theory behind their operation. Moving beyond the digital abstraction, we explore the concepts of bit-streams (sequences of pulses acting as time-based signals) and probabilistic-bits or p-bits (values that can be either 1 or 0 with an assigned probability), as more suitable candidates for the encoding and processing of information in genetic networks. Specifically, the conceptualization of signals as stochastic bit-streams allows for encoding information in the frequency of random expression pulses, offering advantages such as robustness to noise. Additionally, the notion of p-bit enables the design of genetic circuits with capabilities surpassing those of current genetic logic gates, including invertibility. We design several circuits to illustrate these advantages and provide mathematical models and computational simulations that demonstrate their functionality. Our approach to stochastic and probabilistic computing in living cells not only enhances and reflects understanding of information processing in biological systems but also presents promising avenues for designing genetic circuits with advanced functionalities.

synthetic biology↗

bioRxiv2024

Gene regulatory network inference from single-cell data using optimal transport

Modelling gene expression is a central problem in systems biology. Single-cell technologies have revolutionised the field by enabling sequencing at the resolution of individual cells. This results in a much richer data compared to what is obtained by bulk technologies, offering new possibilities and challenges for gene regulatory network inference. Here we introduce GRIT -- a method to fit a differential equation model and to infer gene regulatory networks from single-cell data using the theory of optimal transport. The idea consists in tracking the evolution of the cell distribution over time and finding the system that minimises the transport cost between consecutive time points.

bioinformatics↗