Biology subjects

Scalia, G.

Publications and source records attributed to Scalia, G..

17 recordsLinked to original sources

bioRxiv2026

Virtual phenotypic screening discovers novel scaffolds inhibiting the PI3K/mTOR pathway

Phenotypic drug discovery has yielded many first-in-class small-molecule drugs by discovering modulators of disease phenotypes in physiologically relevant cellular systems. However, high-content phenotypic assays lack the ultra-high-throughput scalability of target-based screens. Recent advances in virtual screening present an opportunity to address this bottleneck, but have been limited to simple phenotypes like viability, restricted to small repurposing libraries, or lack in-depth biological validation. Here, we present PhenoCompass, a multimodal co-embedding model that aligns compound structures and high-content phenotypic imaging to enable virtual phenotypic screening over billion-compound libraries. Following training on the Joint Undertaking in Morphology dataset with more than 100,000 Cell Painting compound profiles, retrospective validation with historical biochemical high-throughput screening data demonstrates that PhenoCompass ranks compounds according to their biochemical target engagement. Leveraging PhenoCompass, we performed a prospective screen of 3.8 billion Enamine REAL compounds for inhibitors of PI3K/mTOR pathway, a critical signaling cascade whose aberrant activation is a common tumor driver. This search identified 11 novel compounds with pathway-consistent Cell Painting readout and diverse scaffolds, a 54-fold enrichment over the training set. Orthogonal validation experiments using a FOXO3A reporter assay and direct kinase inhibition confirmed seven structurally novel inhibitors with distinct mechanisms of action. These results highlight the convergence of diverse molecular target profiles onto a shared morphological pathway signature and establish PhenoCompass as a robust framework for high-content phenotypic virtual screening.

bioinformatics↗

bioRxiv2026

TxConformal: Controlling False Discoveries in AI-Driven Therapeutic Discovery

Artificial Intelligence (AI) is transforming therapeutic discovery by scoring a large set of promising candidates and prioritizing a shortlist for further investigation. Quantifying the reliability of AI scores and preventing false positives among selected candidates is key to the efficiency of the discovery process. Conformal prediction (CP) has emerged as a popular tool for guiding such prioritization, especially via the conformal selection framework to control false discovery rates (FDR) in selecting top-ranked candidates under distributional shift1, 2. However, deploying these advances in real-world therapeutic discovery remains challenging: distribution shifts are difficult to quantify and correct in high-dimensional biomedical data, and practical workflows often require flexible error metrics. Here, we present TO_SCPLOWXC_SCPLOWCO_SCPLOWONFORMALC_SCPLOW, a general framework for trustworthy decision making when building shortlists using AI scores. TO_SCPLOWXC_SCPLOWCO_SCPLOWONFORMALC_SCPLOW adjusts for distribution shift by balancing the hidden representations in AI models and then provides confidence measures for true discoveries of target biological properties. These confidence measures, interpretable as p-values, can be used in conjunction with statistical multiple testing procedures to derive selection decisions with limited false positives or to estimate the errors in given selection decisions. TO_SCPLOWXC_SCPLOWCO_SCPLOWONFORMALC_SCPLOW controls the false positive rate in six real-world tasks spanning various therapeutic discovery stages, modalities, and AI models with realistic data splits. When selecting promising combinatorial genetic perturbations, TO_SCPLOWXC_SCPLOWCO_SCPLOWONFORMALC_SCPLOW nearly halves false-positive selections compared to baseline methods, substantially reducing unnecessary experimental costs by tens of thousands of dollars. When selecting stable protein structures under mutant shifts, TO_SCPLOWXC_SCPLOWCO_SCPLOWONFORMALC_SCPLOW identifies about 10 times more proteins than baseline methods at stringent thresholds when running at a target FDR level of 10%, recovering over 90% of valuable candidates that baseline methods miss due to unaccounted distribution shifts. Furthermore, we demonstrate that TO_SCPLOWXC_SCPLOWCO_SCPLOWONFORMALC_SCPLOW robustly supports various alternative error metrics suitable for resource-constrained settings. Finally, in a prospective fixed-budget virtual screening campaign for novel antibiotic discovery, TO_SCPLOWXC_SCPLOWCO_SCPLOWONFORMALC_SCPLOW predicted false positives in close agreement with experimental outcomes, with substantial improvements over simple baselines.

bioinformatics↗

bioRxiv2026

Label-Free Nucleoli Measurement by 3D Holo-Tomographic Flow Cytometry Using Biolens Phase Compensation

Holo-Tomographic Flow Cytometry (HTFC) holds the potential to transform cellular research and clinical screening through 3D label-free quantitative phase imaging (QPI) of flowing single cells. However, it has been limited by insufficient intracellular specificity in 3D refractive index (RI) distributions, since suspended cells act as highly aberrating spherical biolenses obscuring internal structures. Here, we show the Biolens Phase Compensation (BPC), a method that corrects phase aberrations in 2D QPI projections to transform the 3D RI tomogram. Working within this new 3D pseudo-RI space demonstrates for the first time the extraction of nucleoli in HTFC. Extensive validation against 2D fluorescence flow cytometry and 3D confocal microscopy demonstrates that BPC achieves reliable intranuclear specificity. Using statistically significant single-cell analysis, we provide multiplexed quantitative 3D measurements of nested intracellular compartments (cytoplasm, nucleoplasm, nucleoli). This approach extends label-free HTFC toward capabilities of gold-standard Fluorescence Microscopy, overcoming its well-known drawbacks in intracellular and intranuclear staining.

biophysics↗

bioRxiv2025

Nona: A unifying multimodal masking framework for functional genomics

The non-coding genome encodes complex regulatory logic that orchestrates gene expression and cell identity. While machine learning models for functional genomics have advanced our understanding of the cis-regulatory code, sequence-to-function models, DNA language models, and generative models have evolved as separate paradigms despite probing the same underlying regulatory biology. We introduce Nona, a multimodal masked modeling framework that unifies these paradigms by learning jointly from DNA sequence and base-resolution functional genomics data. Beyond unifying existing modeling paradigms, Nona enables entirely new modeling objectives. We demonstrate its versatility through three applications: (1) a context-aware sequence-to-function model that improves local predictions by up to 13% by correcting systematic errors in sequence-to-function predictions; (2) a functional language model that integrates functional data into language modeling, learns relevant regulatory sequence motifs, and enables regulatory element design through masked discrete diffusion; (3) functional genotyping, which reveals an unrecognized privacy vulnerability in processed ATAC-seq data and re-identifies individuals from genetic databases with perfect accuracy. Together, these results establish masking as a universal interface for integrated modeling of functional genomics data, unifying disparate approaches while opening new directions for understanding and engineering the regulatory genome.

genomics↗

bioRxiv2025

Inferring cellular communication through mapping cells in space using Tangram 2

Cell-to-cell communication (CCC) shapes development, immunity, and disease, yet current spatial transcriptomics (SRT) platforms rarely achieve both single-cell resolution and high-quality transcriptome-wide coverage to accurately characterize CCC in the tissue microenvironment. Tangram2 bridges this gap by integrating single-cell RNA sequencing (scRNA-seq) with SRT to identify genes whose expression changes as a function of neighboring cell types. By accurately mapping cells in tissue space, Tangram2 disentangles interaction-driven transcriptional shifts from intrinsic identity markers, yielding mechanistic CCC maps. Validation across diverse settings--including Slide-tags, co-cultured experiments, human lymph nodes, and simulations--demonstrates high accuracy. Applied to triple-negative breast cancer (TNBC) and cutaneous squamous cell carcinoma (cSCC), Tangram2 recapitulates known biology and uncovers new hypotheses, including various immunosuppressive mechanisms in TNBC and a macrophage-regulatory T-cell circuit associated with survival in cSCC.

genomics↗

bioRxiv2025

Holotomography-driven learning for in-silico staining of single cells in flow cytometry avoiding co-registration

Virtual staining is the current state-of-the-art computational technique to cleverly enhance intracellular specificity in unstained biological samples by using convolutional neural networks (CNNs) trained on co-registered pairs of unstained/stained images. While effective, this approach suffers from unpredictable biases inherent to fluorescence microscopy and encounters challenges when applied to flow cytometry data as it would require accurate co-registration on a huge number of images. Here, we present a novel method that exploits for the first time a Holotomography-driven learning to completely eliminate the need for co-registration. We demonstrate that training a CNN on a stain-free dataset of 3D refractive index tomograms of flowing cells elegantly unlocks stain-free intracellular specificity in quantitative phase imaging flow cytometry. This breakthrough, by circumventing the critical co-registration bottleneck, opens unprecedented perspectives for label-free, high-throughput imaging flow cytometry, offering a powerful new paradigm for advanced 2D and 3D single-cell analysis.

biophysics↗

bioRxiv2025

Deep Evolutionary Fitness Inference for Variant Nomination from Directed Evolution

Iterative screening techniques, such as directed evolution, enable high-throughput affinity maturation to optimize binders to molecular interfaces. However, the decision problem of selecting variants from rich, evolved populations to enter low-throughput follow-up methods remains a significant bottleneck. Here, we present evolutionary fitness inference (EVFI) and DeepEVFI, two machine learning methods that model directed evolution from time-series sequencing data, and infer fitness, a variants ability to enrich under selection pressure. Our methods flexibly handle mutation mechanisms and starting populations that may be partially unknown - settings relevant to drug discovery - and achieve strong performance on a diverse set of experimental data. We conducted two experimental directed evolution campaigns, using antibodies and macrocyclic peptides libraries to identify and optimize binders to therapeutically relevant targets. EVFI and DeepEVFI identified tighter binders that were missed by human experts using conventional frequency-based approaches, including "rising stars" with low frequency. Beyond initial hit discovery, EVFI and Deep-EVFI enables labeling large-scale sequence-fitness datasets and identifying variants of initial binders with diverse properties.

bioinformatics↗

bioRxiv2025

Foundation Model Attributions Reveal Shared Inflammatory Program Across Diseases

Determining a genes functional significance within a cellular context has long been a challenge, as absolute expression level is an unreliable indicator. We introduce SIGnature, a framework for scoring gene importance by leveraging attributions derived from single-cell RNA-sequencing (scRNA-seq) foundation models. Attribution scores reduce technical noise, emphasize regulatory genes, and facilitate cross-dataset comparison - a core challenge for scRNA-seq analyses. We developed the SIGnature package as a tool for generating and querying attributions, enabling rapid gene set searches across massive scRNA-seq atlases. We demonstrated its utility using the MS1 monocyte signature, a poorly understood gene program activated in severe COVID-19 and sepsis. Searching 400 studies revealed novel associations between the MS1 signature and multiple hyperinflammatory conditions, including Kawasaki disease. Experimental validation confirmed Kawasaki disease patient serum induces the MS1 phenotype. These findings highlight that SIGnature can uncover shared mechanisms across conditions, demonstrating its power for large-scale signature scoring and cross-disease analysis.

bioinformatics↗

bioRxiv2025

Gene-embedding-based prediction and functional evaluation of perturbation expression responses with PRESAGE

Understanding the impact of genetic perturbations on cellular behavior is crucial for biological research, but comprehensive experimental mapping remains infeasible. We introduce PRESAGE (Perturbation Response EStimation with Aggregated Gene Embeddings), a simple, modular, and interpretable framework that predicts perturbation-induced expression changes by integrating diverse knowledge sources via gene embeddings. PRESAGE transforms gene embeddings through an attention-based model to predict perturbation expression outcomes. To assess model performance, we introduce a comprehensive evaluation suite with novel functional metrics that move beyond traditional regression tasks, including measures of accuracy in effect size prediction, in identifying perturbations with similar expression profiles (phenocopy), and in prediction of perturbations with the strongest impact on specific gene set scores. PRESAGE outperforms existing methods in both classical regression metrics and our novel functional evaluations. Through ablation studies, we demonstrate that knowledge source selection is more critical for predictive performance than architectural complexity, with cross-system Perturb-seq data providing particularly strong predictive power. We also find that performance saturates quickly with training set size, suggesting that experimental design strategies might benefit from collecting sparse perturbation data across multiple biological systems rather than exhaustive profiling of individual systems. Overall, PRESAGE establishes a robust framework for advancing perturbation response prediction and facilitating the design of targeted biological experiments, significantly improving our ability to predict cellular responses across diverse biological systems.

bioinformatics↗

bioRxiv2025

PerTurboAgent: A Self-Planning Agent for Boosting Sequential Perturb-seq Experiments

Understanding how genetic interventions affect a cells phenotype is key to revealing causal gene regulatory mechanisms and finding novel drug targets. Pooled, high-content perturbation screening methods like Perturb-seq allow us to assess the impact of each of a large number of genetic interventions on a rich cellular profile of RNA or other features, facilitating such discoveries. However, the overall scale of perturbations, especially when considering combinations of genes for perturbation, cannot be tackled exhaustively in the lab. An alternative is to use an iterative design: By leveraging the modularity and sparsity of gene circuits along with prior biological knowledge, we can predict the impact of unseen genetic perturbations on these profiles and group genes with similar effects into co-functional modules, followed by a new round of Perturb-Seq to test these predictions and improve the overall performance of the model. These iterative cycles of experiment and prediction allow prioritizing genes for testing, maximizing the knowledge gleaned from fixed experimental resources, and opening the way to learn general predictive models. Designing these experiments requires a system that can analyze a cellular system, incorporate new and existing knowledge, use statistical tools, predict the effects of unseen perturbations, and prioritize the set of perturbations for the next iteration. These can be time-consuming tasks for scientists and require multiple different skills. Here, we developed PerTurboAgent, an LLM-based agent that excels in predicting candidate gene panels for iterative Perturb-Seq experiments through self-directed data analysis and knowledge retrieval. We evaluated PerTurboAgent based on its ability to identify genes with a phenotypic impact on gene expression upon perturbation in genome-scale perturbation data. PerTurboAgent outperforms existing agent-based and active learning strategies, offering an efficient and understandable approach to designing sequential perturbation experiments.

bioinformatics↗

bioRxiv2025

SpatialAgent: An Autonomous AI Agent for Spatial Biology

Advances in AI are transforming scientific discovery, yet spatial biology, a field that deciphers the molecular organization within tissues, remains constrained by labor-intensive workflows. Here, we present SpatialAgent, an autonomous AI agent for spatial biology research. SpatialAgent couples large language models with a Plan-Act-Conclude architecture, dynamic tool and skill retrieval, multimodal interpretation, and verification modules that audit generated claims. It supports the full discovery loop, from gene-panel design and multimodal annotation to trajectory inference, cell-cell communication analysis, imputation, and hypothesis generation. Across human and mouse brain, heart, tonsil, colon, and prostate datasets, SpatialAgent outperformed established computational baselines and matched or surpassed expert scientists in key tasks. In open-ended case studies, it recovered known tissue organization and generated spatially grounded hypotheses. In a prospective mouse prostate cancer Xenium study, it designed a compact 100-gene add-on panel that profiled 4.2 million cells across 21 samples, improved cell-type and malignant-state prediction, and captured spatially structured tumor and microenvironment programs. SpatialAgent establishes a framework for autonomous and collaborative discovery in spatial biology.

bioinformatics↗

bioRxiv2024

Learning multi-cellular representations of single-cell transcriptomics data enables characterization of patient-level disease states

Single-cell RNA-seq (scRNA-seq) has become a prominent tool for studying human biology and disease. The availability of massive scRNA-seq datasets and advanced machine learning techniques has recently driven the development of single-cell foundation models that provide informative and versatile cell representations based on expression profiles. However, to understand disease states, we need to consider entire tissue ecosystems, simultaneously considering many different interacting cells. Here, we tackle this challenge by generating patient-level representations derived from multi-cellular expression context measured with scRNA-seq of tissues. We develop PaSCient, a novel model that employs a multi-level representation learning paradigm and provides importance scores at the individual cell and gene levels for fine-grained analysis across multiple cell types and gene programs characteristic of a given disease. We apply PaSCient to learn a disease model across a large-scale scRNA-seq atlas of 24.3 million cells from over 5,000 patients. Comprehensive and rigorous benchmarking demonstrates the superiority of PaSCient in disease classification and its multiple downstream applications, including dimensionality reduction, gene/cell type prioritization, and patient subgroup discovery.

bioinformatics↗

bioRxiv2024

Decoding sequence determinants of gene expression in diverse cellular and disease states

Sequence-to-function models that predict gene expression from genomic DNA sequence have proven valuable for many biological tasks, including understanding cis-regulatory syntax and interpreting non-coding genetic variants. However, current state-of-the-art models have been trained largely on bulk expression profiles from healthy tissues or cell lines, and have not learned the properties of precise cell types and states that are captured in large-scale single-cell transcriptomic datasets. Thus, they lack the ability to perform these tasks at the resolution of specific cell types or states across diverse tissue and disease contexts. To address this gap, we present Decima, a model that predicts the cell type- and condition- specific expression of a gene from its surrounding DNA sequence. Decima is trained on single-cell or single-nucleus RNA sequencing data from over 22 million cells, and successfully predicts the cell type-specific expression of unseen genes based on their sequence alone. Here, we demonstrate Decimas ability to reveal the cis-regulatory mechanisms driving cell type-specific gene expression and its changes in disease, to predict non-coding variant effects at cell type resolution, and to design regulatory DNA elements with precisely tuned, context-specific functions.

genomics↗

bioRxiv2024

A high-throughput phenotypic screen combined with an ultra-large-scale deep learning-based virtual screening reveals novel scaffolds of antibacterial compounds

The proliferation of multi-drug-resistant bacteria underscores an urgent need for novel antibiotics. Traditional discovery methods face challenges due to limited chemical diversity, high costs, and difficulties in identifying structurally novel compounds. Here, we explore the integration of small molecule high-throughput screening with a deep learning-based virtual screening approach to uncover new antibacterial compounds. Leveraging a diverse library of nearly 2 million small molecules, we conducted comprehensive phenotypic screening against a sensitized Escherichia coli strain that, at a low hit rate, yielded thousands of hits. We trained a deep learning model, GNEprop, to predict antibacterial activity, ensuring robustness through out-of-distribution generalization techniques. Virtual screening of over 1.4 billion compounds identified potential candidates, of which 82 exhibited antibacterial activity, illustrating a 90X improved hit rate over the high-throughput screening experiment GNEprop was trained on. Importantly, a significant portion of these newly identified compounds exhibited high dissimilarity to known antibiotics, indicating promising avenues for further exploration in antibiotic discovery.

bioinformatics↗

bioRxiv2023

Scalable querying of human cell atlases via a foundational model reveals commonalities across fibrosis-associated macrophages

Single-cell RNA-seq (scRNA-seq) studies have profiled over 100 million human cells across diseases, developmental stages, and perturbations to date. A singular view of this vast and growing expression landscape could help reveal novel associations between cell states and diseases, discover cell states in unexpected tissue contexts, and relate in vivo cells to in vitro models. However, these require a common, scalable representation of cell profiles from across the body, a general measure of their similarity, and an efficient way to query these data. Here, we present SCimilarity, a metric learning framework to learn and search a unified and interpretable representation that annotates cell types and instantaneously queries for a cell state across tens of millions of profiles. We demonstrate SCimilarity on a 22.7 million cell corpus assembled across 399 published scRNA-seq studies, showing accurate integration, annotation and querying. We experimentally validated SCimilarity by querying across tissues for a macrophage subset originally identified in interstitial lung disease, and showing that cells with similar profiles are found in other fibrotic diseases, tissues, and a 3D hydrogel system, which we then repurposed to yield this cell state in vitro. SCimilarity serves as a foundational model for single cell gene expression data and enables researchers to query for similar cellular states across the entire human body, providing a powerful tool for generating novel biological insights from the growing Human Cell Atlas.

bioinformatics↗

bioRxiv2022

Reference-based cell type matching of spatial transcriptomics data

With the advent of multiplex fluorescence in situ hybridization (FISH) and in situ RNA sequencing technologies, spatial transcriptomics analysis is advancing rapidly. Spatial transcriptomics provides spatial location and pattern information about cells in tissue sections at single cell resolution. Cell type classification of spatially-resolved cells can also be inferred by matching the spatial transcriptomics data to reference single cell RNA-sequencing (scRNA-seq) data with cell types determined by their gene expression profiles. However, robust cell type matching of the spatial cells is challenging due to the intrinsic differences in resolution between the spatial and scRNA-seq data. In this study, we systematically evaluated six computational algorithms for cell type matching across four spatial transcriptomics experimental protocols (MERFISH, smFISH, BaristaSeq, and ExSeq) conducted on the same mouse primary visual cortex (VISp) brain region. We find that while matching results of individual algorithms vary to some degree, they also show agreement to some extent. We present two ensembl meta-analysis strategies to combine the individual matching results and share the consensus matching results in the Cytosplore Viewer (https://viewer.cytosplore.org) for interactive visualization and data exploration. The consensus matching can also guide spot-based spatial data analysis using SSAM, allowing segmentation-free cell type assignment.

bioinformatics↗

bioRxiv2020

Deep learning and alignment of spatially-resolved whole transcriptomes of single cells in the mouse brain with Tangram

Charting a biological atlas of an organ, such as the brain, requires us to spatially-resolve whole transcriptomes of single cells, and to relate such cellular features to the histological and anatomical scales. Single-cell and single-nucleus RNA-Seq (sc/snRNA-seq) can map cells comprehensively5,6, but relating those to their histological and anatomical positions in the context of an organs common coordinate framework remains a major challenge and barrier to the construction of a cell atlas7-10. Conversely, Spatial Transcriptomics allows for in-situ measurements11-13 at the histological level, but at lower spatial resolution and with limited sensitivity. Targeted in situ technologies1-3 solve both issues, but are limited in gene throughput which impedes profiling of the entire transcriptome. Finally, as samples are collected for profiling, their registration to anatomical atlases often require human supervision, which is a major obstacle to build pipelines at scale. Here, we demonstrate spatial mapping of cells, histology, and anatomy in the somatomotor area and the visual area of the healthy adult mouse brain. We devise Tangram, a method that aligns snRNA-seq data to various forms of spatial data collected from the same brain region, including MERFISH1, STARmap2, smFISH3, and Spatial Transcriptomics4 (Visium), as well as histological images and public atlases. Tangram can map any type of sc/snRNA-seq data, including multi-modal data such as SHARE-seq data5, which we used to reveal spatial patterns of chromatin accessibility. We equipped Tangram with a deep learning computer vision pipeline, which allows for automatic identification of anatomical annotations on histological images of mouse brain. By doing so, Tangram reconstructs a genome-wide, anatomically-integrated, spatial map of the visual and somatomotor area with [~]30,000 genes at single-cell resolution, revealing spatial gene expression and chromatin accessibility patterning beyond current limitation of in-situ technologies.

genomics↗