Biology subjects

Elemento, O.

Publications and source records attributed to Elemento, O..

At least 19 recordsLinked to original sources

bioRxiv2026

Constrained Generative Design Frameworks For Computational Discovery of Target-Specific DARPin Candidates

Applying unconstrained generative protein models to fixed structural scaffolds can produce systematic design artifacts, including a "Glycine Trap" characterized by the enrichment of glycine at structurally incompatible positions. Furthermore, optimizing sequences against artificial rigid-body docking geometries induces reward-hacking and severe geometric hallucinations. In addition, the highly conserved designed ankyrin repeat protein, or DARPin, scaffold can obscure defects at the engineered binding interface, causing AlphaFold2-Multimer (AF2) to predict nonfunctional protein-target interactions with high confidence. To overcome these limitations, we developed DARPinMPNN, a scaffold-constrained computational pipeline for DARPin candidate discovery. Restricting sequence generation to a validated DARPin design space eliminated these failure modes. A state-aware chimeric multiple sequence alignment strategy was engineered and enabled AlphaFold2-Multimer (AF2) to serve as a high-throughput structural sieve, while AlphaFold 3 (AF3) provided independent structural validation of candidate binders. Using this framework, we identified mesothelin-targeting DARPin candidates with predicted structural confidences (champion ipTM = 0.83) approaching those of a structurally validated picomolar-affinity binder (G3 control, ipTM = 0.89). By revealing extensive discordance between AF2 and AF3 predictions, this work establishes a robust framework for identifying and prioritizing high-confidence DARPin candidates for experimental validation.

bioinformatics↗

bioRxiv2026

Spatial Proteomics Reveals Dual-Compartment Immune Evasion Architecture and Core Lymphomagenic Stromal Niches in Follicular Lymphoma

Follicular lymphomas (FL) depend on microenvironmental signals for survival, and the composition of the tumor microenvironment may influence clinical outcome. How it is spatially organized, and how that organization drives immune evasion, has not been established. We applied two complementary 39-marker imaging mass cytometry panels to 156 follicular lymphoma tumor cores, generating a spatial atlas of 4.2 million cells resolved into nine tissue compartments. FL follicles operated as immune-privileged zones: CD8 T cells were largely excluded, and those that did infiltrate were exhausted in a gradient that peaked at the GC core and dissolved at the follicle boundary. A Treg-enriched barrier at that boundary separated the intrafollicular program from an effector-rich interfollicular compartment. Within follicles, CD14-high follicular dendritic cells defined an activated stromal niche that supported tumor B cell proliferation and predicted shorter progression-free and overall survival. Across patients, per-cell CD14 intensity was the strongest predictor of progression, remaining prognostic independently of FLIPI, grade, and stage. VISTA, not PD-L1, dominated checkpoint expression across all myeloid subtypes in both follicular and interfollicular zones, and was further amplified on M2 macrophages in patients carrying EP300 mutations. S100A9+ MDSC-like cells were the strongest myeloid predictor of histologic transformation. Agent-based modeling of this architecture predicted synergistic tumor clearance when anti-VISTA was combined with a CD20-CD3 bispecific antibody. These findings establish that FL uses tissue architecture as an immune evasion mechanism and identify VISTA and CD14-high follicular dendritic cells as candidate therapeutic targets.

Cancer Biology↗

bioRxiv2026

A Foundation Model for the Cancer Genome

AO_SCPLOWBSTRACTC_SCPLOWCancer is a disease of the genome, in which somatic mutations and copy-number alterations determine tumour identity, clinical behaviour, and response to therapy. Consortium-scale sequencing has profiled hundreds of thousands of tumours,1,2 yet clinical interpretation still proceeds one alteration at a time against hand-curated knowledgebases,3,4 often ignoring co-occurring alterations and the genome-wide copy-number pattern. Self-supervised foundation models pretrained on unlabelled corpora5 have produced transferable representations in adjacent biological domains6-8 by learning joint structure across many features, yet no comparable model exists for the cancer genome. Here we present TESSERA (Tumour Embeddings via Self-Supervised Encoding and Reconstruction of Alterations), a foundation model for the cancer genome; we pretrain it on somatic single-nucleotide variants and copy-number segments through masked-token reconstruction within each modality and a contrastive objective across modalities. A single representation, produced once and reused without retraining, supports variant pathogenicity prediction, pan-cancer tumour typing, unsupervised molecular subtyping, prognostic stratification, and counterfactual treatment-effect estimation that yields predictive chemotherapy-selection biomarkers in real-world cohorts. These biomarkers are interpretable: each surfaces the co-occurring alterations underlying the prediction, exposing biology that single-gene rules miss. In metastatic colorectal cancer, where the FOLFOX-vs-FOLFIRI choice is currently guided by toxicity rather than tumour biology, the model uncovers a candidate predictive biomarker: a three-feature rule (TP53+/KRAS+/17p-) selecting patients who derive substantially greater benefit from FOLFOX than FOLFIRI.

bioinformatics↗

bioRxiv2026

Mapping human natural killer cell development in tonsil

Secondary lymphoid tissue, including tonsil, supports human NK cell development, but the spatial organization and tissue niches that drive this differentiation remain undefined. Here, we used single cell analysis of cyclic immunofluorescence to generate a comprehensive atlas of human NK cell development in tonsil. By integrating regional localization, chemokine signaling, cytokine availability, and cell phenotype, we show that NK cell differentiation follows a reproducible spatial trajectory defined by stage-specific cell-cell interactions. CD34+ NK cell progenitors are found in the interfollicular domain in proximity to high endothelial venules, consistent with this route of progenitor entry into tissue. Mature NK cells are primarily found in the T cell-rich parafollicular domain, where they interact with other NK cells and T cell subsets. Local inflammation increases NK cell frequency in tissue through both proliferation of NK progenitors and increased frequency of mature NK cells. Finally, we identify a subset of tonsil stromal cells that support differentiation of NK cells in vitro and proliferation of NK precursors in situ. Together, these findings demonstrate that spatial localization defines human NK cell development and provide an in situ definition of niches that support human NK cell differentiation in tonsil.

immunology↗

bioRxiv2026

Mutant p53 Directs PARP to Regulate Replication Stress and Drive Breast Cancer Metastasis

TP53 mutations occur in 80-90% of triple-negative breast cancers (TNBCs) and drive genomic instability and metastatic progression. Poly (ADP-ribose) polymerase (PARP) is critical for DNA repair and replication fork stability. How oncogenic signaling influences PARP function to sustain proliferation during replication stress remains unclear. Mutant p53 (mtp53) R273H associates tightly with chromatin, forms complexes with PARP, and enhances PARP recruitment to replication forks [1-3]. The C-terminal region of mtp53 mediates mtp53-PARP and mtp53-Poly (ADP-ribose) (PAR) interactions that facilitate S phase progression [4, 5]. The PARP inhibitor talazoparib (TAL) combined with the alkylating agent temozolomide (TMZ) produces synergistic cytotoxicity selectively in mtp53, but not wild-type p53 (wtp53), breast cancer cells and organoids. Herein we evaluated the mechanism of mtp53-associated cell death and tested if this could translate to a preclinical xenograft model. We found that TMZ+TAL treatment induced elevated cleaved PARP and {gamma}H2AX and reduced the metastasis-promoting oncoprotein MDMX. In orthotopic xenografts expressing mtp53 R273H, but not wtp53, combination therapy significantly decreased circulating tumor cells (CTCs) and lung metastases. Transcriptomic profiling of tumors from combination treated animals demonstrated downregulation of MDMX, VEGF, and NF-{kappa}B, consistent with the observed suppression of CTCs and lung metastasis, and increased {gamma}H2AX, indicative of replication stress in mtp53 xenografts. Inhibition of metastasis was also observed in mtp53 R273H WHIM25 and p53-undetectable WHIM6 TNBC patient-derived xenografts (PDX). The mtp53 C-terminal domain (347-393) demonstrated a critical tumor promoting function, as CRISPR-mediated deletion impaired replication fork progression, tumor growth, and metastatic dissemination. DNA fiber combing showed that expression of full-length mtp53 R273H, but not C-terminal deleted {Delta}347-393, supported sustained single-stranded DNA gaps (ssGAPs) following Poly (ADP-ribose) glycohydrolase (PARG) inhibition. These findings support that mtp53 uses C-terminal amino acids to exploit PARP to enable replication stress adaptation and that mtp53 is a predictive biomarker for combined PARP inhibitor and DNA damaging therapies targeting TNBC. Significance statementTP53 mutations are the most common genetic alterations in TNBC and a major driver of replication stress and metastasis. This study shows that missense mutant p53 uses C-terminal amino acids to reprogram PARP activity to maintain tumor cell survival under replication stress. We demonstrate that p53 status governs the response to combined PARP inhibitor (PARPi) and DNA-damaging chemotherapy, establishing an additional molecular basis beyond BRCA1 mutations for treating TNBC with PARPi therapy. These findings reveal a previously unrecognized mechanism by which the mutant p53-PARP axis enables replication stress tolerance and drives cancer metastasis. We show mutation of p53 in TNBC provides an additional biomarker-guided framework to improve PARPi therapeutic outcomes.

cancer biology↗

bioRxiv2026

BEACON: predicting side effects and therapeutics outcomes to drugs by Bridging knowlEdge grAph with CONtextual language model

Biomedical knowledge graphs encode millions of relationships between drugs, proteins, pathways, and diseases, yet translating this structured knowledge into accurate predictions remains challenging. Existing deep learning approaches, including graph neural networks and knowledge graph embeddings, assign fixed representations to entities regardless of biological context, limiting their ability to capture how the same gene or pathway functions differently across scenarios. These methods also lack interpretability and often fail when applied to novel drugs outside their training distribution. Here we present BEACON (Bridging knowlEdge grAph with CONtextual language model), a framework that transforms knowledge graphs into contextual sentence representations processable by language models. BEACON converts biomedical entities into tokens and relationships into syntactic dependencies, creating "sentence trees" that preserve graph structure while enabling contextual processing. A visibility matrix ensures that attention patterns respect the underlying knowledge graph topology, and a perturbation-based evaluation module identifies the specific genes, enzymes, and pathways driving each prediction. We demonstrate BEACONs versatility through two clinically important applications. For drug sensitivity prediction in cancer cell lines, BEACON achieves 0.941 AUROC and Spearman {rho} = 0.919, outperforming existing methods (DrugCell, DeepCDR and DeepTTA). For drug-drug interaction (DDI) prediction, BEACON achieves 0.964 AUROC on the TwoSIDES benchmark and 0.84 AUROC on temporally held-out FDA adverse event data (2013-2023), demonstrating robust generalization to newly approved drugs. Applying BEACON to the BTK inhibitor acalabrutinib revealed that predicted interactions are enriched for drugs metabolized by CYP3A enzymes (OR = 3.01, P = 4.3 x 10-), a mechanism validated through network proximity analysis. BEACON provides a unified, interpretable approach to knowledge graph-enhanced biomedical prediction.

bioinformatics↗

bioRxiv2025

APOBEC3A-Induced DNA Damage Drives Polymerase Theta Dependency and Synthetic Lethality in Cancer

APOBEC3 cytidine deaminases drive cancer evolution. There is an unmet need to target cancer cells with APOBEC3 activity. Here, we identify error-prone theta-mediated end joining (TMEJ) as the main pathway for repairing APOBEC3-induced double-strand breaks (DSBs). Using fluorescent DSB repair reporters and a novel biochemical assay, we demonstrate that APOBEC3A competes with replication protein A (RPA) for single-stranded DNA overhangs, exposing microhomologous sequences to shift DSB repair towards error-prone TMEJ. Genomic analysis of clinical tumor samples confirmed the co-occurrence and proximity between APOBEC3-induced mutational footprints, microhomology-mediated deletions (MMDs), and TMEJ-associated chromosomal instability signatures. Crucially, inhibition of DNA polymerase theta (Pol{theta}) synergizes with APOBEC3A-induced DSBs to induce synthetic lethality in vitro and in vivo. Collectively, our findings identify TMEJ as the preferred mechanism for repairing APOBEC3A-induced DSBs and establish Pol{theta} inhibition as a novel promising strategy to eliminate cancer cells with APOBEC3A activity.

cancer biology↗

bioRxiv2025

Tumor-draining lymph node derived CD8⁺ T cells sustain durable systemic immunity after neoadjuvant IL-15 and PD-1 blockade.

The tumor-draining lymph node (tdLN) serves as a critical reservoir of tumor-reactive T cells, yet its contribution to preventing metastatic recurrence remains poorly understood. In this study, we investigate CD8 T cell clonal dynamics and memory responses induced by neoadjuvant anti-PD-1 therapy. We demonstrate that PD-1 blockade promotes the emergence of protective CD8 T cell memory, enriched with tumor-relevant clones that persist in the lymph nodes of long-term survivors. Furthermore, we show that combination of a neoadjuvant IL-15 superagonist with PD-1 blockade significantly increases clonal diversity, cytotoxicity, and clonal persistence in non-draining lymph nodes. These changes result in enhanced systemic antitumor immunity and durable memory that prevents tumor relapse. Our findings establish the tdLN as a critical immunologic hub that can be therapeutically targeted by IL-15 to augment PD-1 blockade efficacy and support a rationale for a phase II neoadjuvant combination immunotherapy trial in early-stage NSCLC. One sentence summaryNeoadjuvant IL-15 with PD-1 blockade expands lymph node clonal diversity and preserves stem-like CD8 T cell features, yielding superior antitumor efficacy, durable memory, and reduced metastatic recurrence.

immunology↗

bioRxiv2025

Paired plus-minus sequencing is an ultra-high throughput and accurate method for dual strand sequencing of DNA molecules

Distinguishing real biological variation in the form of single-nucleotide variants (SNVs) from errors is a major challenge for genome sequencing technologies. This is particularly true in settings where SNVs are at low frequency such as cancer detection through liquid biopsy, or human somatic mosaicism. State-of-the-art molecular denoising approaches for DNA sequencing rely on duplex sequencing, where both strands of a single DNA molecule are sequenced to discern true variants from errors arising from single stranded DNA damage. However, such duplex approaches typically require massive over-sequencing to overcome low capture rates of duplex molecules. To address these challenges, we introduce paired plus-minus sequencing (ppmSeq) technology, in which both DNA strands are partitioned and clonally amplified on sequencing beads through emulsion PCR. In this reaction, both strands of a double-stranded DNA molecule contribute to a single sequencing read, allowing for a duplex yield that scales linearly with sequencing coverage across a wide range of inputs (1.8-98 ng). We benchmarked ppmSeq against current duplex sequencing technologies, demonstrating superior duplex recovery with ppmSeq, with a rate of 44%{+/-}5.5% (compared to [~]5-11% for leading duplex technologies). Using both genomic as well as cell-free DNA, we established error rates for ppmSeq, which had residual SNV detection error rates as low as 7.98x10-8 for gDNA (using an end-repair protocol with dideoxy nucleotides) and 3.5x10-7{+/-}7.5x10-8 for cell-free DNA. To test the capabilities of ppmSeq for error-corrected whole-genome sequencing (WGS) for clinical application, we assessed circulating tumor DNA (ctDNA) detection for disease monitoring in cancer patients. We demonstrated that ppmSeq enables powerful tumor-informed ctDNA detection at concentrations of 10-4 across most cancers, parts per million sensitivity in cancers with high mutation burden, and further increased sensitivity with higher sequencing depth. We then leveraged genome-wide trinucleotide mutation patterns characteristic of urothelial (APOBEC3-related and platinum exposure-related signatures) and lung (tobacco-exposure-related signatures) cancers to perform tumor-naive ctDNA detection, showing that ppmSeq can identify a disease-specific signal in plasma cell-free DNA without a matched tumor, and that this signal correlates with imaging-based disease metrics. Altogether, ppmSeq provides an error-corrected, cost-efficient and scalable approach for high-fidelity WGS that can be harnessed for challenging clinical applications and emerging frontiers in human somatic genetics where high accuracy is required for mutation identification.

genomics↗

bioRxiv2025

Transcriptome sequencing of Hodgkin lymphoma Hodgkin and Reed-Stenberg cells reveals escape from NK cell recognition and an unfolded protein response

The mutational profile of classic Hodgkin lymphoma (cHL) overlaps with that of related B cell lymphomas, including primary mediastinal B cell lymphoma (PMBL), and yet these are different histologically and clinically. To discover the molecular features that distinguish cHL, we deployed flow cytometric cell sorting and low-input RNA sequencing to generate full transcriptome data from viable, isolated Hodgkin and Red-Sternberg (HRS) cells from eighteen primary tumors, alongside matched intra-tumoral non-neoplastic B cells and four cell lines. Comparison of HRS cells to normal cellular subsets revealed evidence of abortive plasma cell differentiation, with an unfolded protein response signature, shared with plasma cell neoplasms, but not other B-cell lymphoma types. Comparison of cHL to PMBL revealed similarities but also key differences in B cell differentiation programs accompanied by upregulation of genes involved in microtubule cytoskeleton organization in cHL, which may be related to the unique multinucleated nature of HRS cells. In HRS cells, we also observed a downregulation of SLAM family receptors, which are crucial for NK cell activation, providing a potential mechanism for immune evasion from NK-mediated killing. STATEMENT OF SIGNIFICANCEThis study defines a unique transcriptional program in classic Hodgkin Lymphoma (cHL) marked by oncogenic signaling, chromatin integrity, DNA repair, and immune escape including loss of NK-activating receptors. HRS cells resemble plasma cells and have an unfolded protein response signature, which distinguishes them from diffuse large and primary mediastinal B cell lymphomas.

cancer biology↗

bioRxiv2025

The Spatial Atlas of Human Anatomy (SAHA): A Multimodal Subcellular-Resolution Reference Across Human Organs

The Spatial Atlas of Human Anatomy (SAHA) represents the first multimodal, subcellular- resolution reference of healthy adult human tissues across multiple organ systems. Integrating spatial transcriptomics, proteomics, and histological features across over 15 million cells from more than 100 donors, SAHA maps conserved and organ-specific cellular niches in gastrointestinal and immune tissues. High-resolution profiling using CosMx SMI, 10x Xenium, RNAscope, GeoMx DSP, and single-nucleus RNA-seq reveals spatially organized cell states, rare adaptive immune populations, and tissue-specific cell-cell interactions and ligand-receptor pairs. Comparative analyses with colorectal cancer and inflammatory bowel disease demonstrate the power of SAHA to detect disease-associated spatial disruptions, including crypt dedifferentiation, perineural invasion, and therapy-resistant immune remodeling. All data are openly accessible through a FAIR-compliant interactive portal to support exploration, benchmarking, and machine learning model training. Through SAHA, we provide a foundational framework for spatial diagnostics and next-generation precision medicine grounded in a comprehensive human tissue atlas, enabling the development of context-aware models that simulate tissue behavior, decode complex pathologies, and accelerate therapeutic innovation at unprecedented scale.

systems biology↗

bioRxiv2025

Tumor-specific draining lymph node CD8 T cells orchestrate an anti-tumor response to neoadjuvant PD-1 immune checkpoint blockade

Elucidating the anti-tumor role of tumor-draining lymph nodes (tdLNs) in patients could offer critical mechanistic insight and shift therapeutic strategies from a tumor-centric approach to one that considers tumor-immune system interplay. Our study characterizes benign tdLNs T cell anti-tumor responses beyond initial T cell priming in patients with resectable non-small cell lung cancer. We further investigated whether tumor-specific tdLN T cells were altered by immune checkpoint blockade (ICB) locally and systemically. We performed single-cell TCR lineage tracing and transcriptomic profiling on 672,886 CD8 T cells from 41 tumor, benign tdLN, and blood samples in 14 patients treated with or without neoadjuvant chemoimmunotherapy (ChemoIO). Using deep-integrating clonal tracking with machine learning-based transcriptional analysis, our findings revealed that benign tdLNs locally and independently orchestrate two transcriptionally distinct tumor-specific memory CD8 T cell populations: one with ZNF683+ CXCR6+ tumor tissue-residency potential, and another with cytotoxic memory potential. Furthermore, tdLN-derived clones not only constitute the dominant tumor-infiltrating (75%) and circulating (>90%) tumor-specific expanded T cell populations but also preserve their transcriptionally distinct subset identities within the tumor T cell effector state. ChemoIO selectively increased the clonal diversity and cytotoxic memory/TEMRA programs of tdLN-derived clones locally and systemically, both of which remained unchanged in clones lacking tdLN TCR lineage. In conclusion, the tdLN locally orchestrates tumor-reactive and ChemoIO-reactive transcriptional distinct T cell subsets that shape the circulating blood and tumor T cell environments. These findings represent a clinical paradigm shift with implications regarding the extent of tdLN resection during surgery, timing of ChemoIO treatment, and the development of memory T cell-based immunotherapies.

cancer biology↗

bioRxiv2025

Simultaneous immunomodulation and epithelial-to-mesenchymal transition drives lung adenocarcinoma progression

Lung cancer remains the deadliest cancer in the United States, with lung adenocarcinoma (LUAD) as its most prevalent subtype. While computed tomography (CT)-based screening has improved early detection and enabled curative surgeries, the molecular and cellular dynamics driving early-stage LUAD progression remain poorly understood, limiting non-surgical treatment options. To address this gap, we profiled 2.24 million cells from 122 early-stage LUAD patients using multiplexed imaging mass cytometry (IMC). This analysis revealed the molecular, spatial, and temporal dynamics of LUAD development. Our findings uncover a binary progression model. LUAD advances through either inflammation, driven by a balance of cytotoxic and regulatory immune activity, or fibrosis, characterized by stromal activation. Surprisingly, tumor cell populations did not increase significantly. Instead, they displayed a mixed phenotypic profile consistent with epithelial-to-mesenchymal transition (EMT), effectively masking the expansion of malignant cells. Furthermore, we addressed discrepancies between CT-based and histology-based subtyping. CT scans, while non-invasive, often mischaracterize invasive fibrotic tumors--which account for 20.5% of LUAD cases--as mild, non-solid ground glass opacities (GGOs). Using high-content IMC imaging, we demonstrate that these tumors harbor significant risks and advocate for improved diagnostic strategies. These strategies should integrate molecular profiling to refine patient stratification and therapeutic decision-making. Altogether, our study provides a high-resolution, systems-level view of the tumor microenvironment in early-stage LUAD. We characterize key transitions in oncogenesis and propose a precision-driven framework to enhance the detection and management of aggressive disease subtypes.

cancer biology↗

bioRxiv2025

Acquisition of discrete immune suppressive barriers contributes to the initiation and progression of preinvasive to invasive human lung cancer.

Computerized chest tomography (CT)-guided screening in populations at risk for lung cancer has increased the detection of preinvasive subsolid nodules, which progress to solid invasive adenocarcinoma. Despite the clinical significance, there is a lack of effective therapies for intercepting the progression of preinvasive to invasive adenocarcinoma. To uncover determinants of early disease emergence and progression, we used integrated single-cell approaches, including scRNA-seq, multiplexed imaging mass cytometry and spatial transcriptomics, to construct the first high-resolution map of the composition, lineage/functional states, developmental trajectories and multicellular crosstalk networks from microdissected non-solid (preinvasive) and solid compartments (invasive) of individual part-solid nodules. We found that early disease initiation and subsequent progression are associated with the evolution of immune-suppressive cellular phenotypes characterized by decreased cytotoxic CD8 T and NK cells, increased T cell exhaustion and accumulation of immunosuppressive regulatory T cells (Tregs) and M2-like macrophages expressing TREM2. Within Tregs, we identified a unique population of 4-1BB+ Treg subset enriched for the IL2-STAT5 suppressive pathway with transcription profiles supporting discrete metabolic alterations. Spatial analysis showed increased density of suppressive immune cells around tumor cells, increased exhaustion phenotype of both CD4 and CD8 T cells expressing chemokine CXCL13, and spatial micro-complex of endothelial and lymphocyte interactions within tertiary lymphoid structures. The single-cell architecture identifies determinants of early disease emergence and progression, which may be developed not only as diagnostic/prognostic biomarkers but also as targets for disease interception. Additionally, our dataset constitutes a valuable resource for the preinvasive lung cancer research community.

cancer biology↗

bioRxiv2024

DDI-GPT: Explainable Prediction of Drug-Drug Interactions using Large Language Models enhanced with Knowledge Graphs

Identifying potential drug-drug interactions (DDIs) before clinical use is essential for patient safety yet remains a significant challenge in drug development. We presented DDI-GPT, a deep learning framework that predicts DDIs by combining knowledge graphs (KGs) and pre-trained large language models (LLMs), enabling early detection of potential drug interactions. We demonstrated that DDI-GPT outperforms current state-of-the-art methods by capturing contextual dependencies between biomedical entities to infer potential DDIs. Through feature attribution methods, we show that our explainable deep learning (DL) models enhance the quality of explanations on the pathways and interactome networks. Using TwoSIDES as our benchmark dataset, DDI-GPT achieved the best performance of 0.964 in AUROC compared with existing DL methods. We also applied DDI-GPT to perform zero-shot prediction on 9,480 DDI records, encompassing 442 distinct drugs from the FDA Adverse Event Reporting System. DDI-GPT can attain a high accuracy of in 0.84 AUROC, with an improvement of 14% compared to the best previously published method. We explored model interpretations on predicted DDIs involving Brutons tyrosine kinase (BTK) inhibitors and uncovered CYP3A-enriched signals underlying the contaminant use of BTK inhibitors with other drugs leading to toxicity. Altogether, DDI-GPT, implemented as both a web server platform and a software package, identifies DDI events and offers a deep learning tool for drug safety use with expandable features.

bioinformatics↗

bioRxiv2024

PAIRWISE: Deep Learning-based Prediction of Effective Personalized Drug Combinations in Cancer

Combination therapies offer promise for improving cancer treatment efficacy and preventing recurrence. However, identifying optimal drug combinations tailored to specific cancer subtypes and individual patients is extremely challenging due to the vast number of possible combinations and tumor heterogeneity. To address this gap, we take a machine learning approach combining deep learning with transfer learning to incorporate prior scientific knowledge and predict drug synergy based on tumor-specific transcriptome profiles. This approach, called PAIRWISE, explicitly modeled synergistic effects of drug combinations in cancer cell lines or individual tumor samples based on drug chemical structures, drug targets, and transcriptomes of inferred samples. PAIRWISE outperformed competing models with an area under the receiver operating characteristic curve (AUROC) of 0.85 on held-out cancer cell lines. When applied to an independent dataset of combinations with Bruton Tyrosine Kinase inhibitors (BTKi) in Diffuse Large B Cell Lymphoma (DLBCL) cell lines, PAIRWISE accurately predicted synergistic drug combinations with an AUROC of 0.72. To further confirm the robustness of PAIRWISE predictions, we performed an in silico, patient profile-directed screen for other compounds that would synergize with BTKi in DLBCL patients, and confirmed the synergy of the predictions using a panel of eight non-Hodgkin lymphoma cell lines. These findings demonstrate the ability of PAIRWISE to nominate effective personalized drug combinations, accelerating the development of precision oncology.

bioinformatics↗

bioRxiv2024

Mechanism of neurodegeneration mediated by clonal inflammatory microglia

Langerhans cell Histiocytosis (LCH) and Erdheim-Chester disease (ECD) are clonal myeloid disorders, associated with MAP-Kinase activating mutations and an increased risk of neurodegeneration. Surprisingly, we found pervasive PU.1+ microglia mutant clones across the brain of LCH and ECD patients with and without neurological symptoms, associated with microgliosis, reactive astrocytosis, and neuronal loss. The disease predominated in the grey nuclei of the rhombencephalon, a topography attributable to a local proliferative advantage of mutant microglia. Presence of clinical symptoms was associated with a longer evolution of the disease and a larger size of PU.1+ clones (p= 0.0003). Genetic lineage tracing of PU.1+ clones suggest a resident macrophage lineage or a bone marrow precursor origin depending on patients. Finally, a CSF1R-inhibitor depleted mutant microglia and limited neuronal loss in mice suggesting an alternative to MAPK inhibitors. These studies characterize a progressive neurodegenerative disease, caused by clonal proliferation of inflammatory microglia (CPIM), with a decade(s)-long preclinical stage of incipient disease that represent a therapeutic window for prevention of neuronal death.

immunology↗

bioRxiv2024

The Orphan G Protein-Coupled Receptor GPR52 is a Novel Regulator of Breast Cancer Multicellular Organization

Statement of SignificanceWe showed that loss of the orphan G protein-coupled receptor GPR52 in human breast cell lines leads to increased cell clustering, hybrid/partial EMT, and increased tumor burden in zebrafish. BackgroundG protein-coupled receptors (GPCRs) are the largest class of membrane-bound receptors that transmit critical signals from extracellular to intracellular spaces. Transcriptomic data of resected breast tumors show that low mRNA expression of orphan GPCR GPR52 correlates with reduced overall survival in patients with breast cancer, leading to the hypothesis that loss of GPR52 supports breast cancer progression. MethodsCRISPR-Cas9 was used to knockout GPR52 in the human triple-negative breast cancer (TNBC) cell lines MDA-MB-468 and MDA-MB-231, and in the non-cancerous breast epithelial cell line MCF10A. 2D and 3D in vitro studies, electron microscopy, Matrigel culture, and a zebrafish xenograft model were used to assess the morphology and behavior of GPR52 KO cells. RNA-sequencing and proteomic analyses were also conducted on these cell lines, and transcriptomic data from The Cancer Genome Atlas (TCGA) database were used to compare GPR52-null and wild-type (WT) signatures in breast cancer. ResultsLoss of GPR52 was found to be associated with increased cell-cell interaction in 2D cultures, altered 3D spheroid morphology, and increased propensity to organize and invade collectively in Matrigel. Furthermore, GPR52 loss was associated with features of EMT in MDA-MB-468 cells, and zebrafish injected with GPR52 KO cells developed a greater total cancer area than those injected with control cells. RNA sequencing and proteomic analyses of GPR52-null breast cancer cells revealed an increased cAMP signaling signature. Consistently, we found that treatment of wild-type (WT) cells with forskolin, which stimulates the production of cAMP, induces phenotypic changes associated with GPR52 loss, and inhibition of cAMP production rescued some GPR52 KO phenotypes. ConclusionGPR52 is an orphan GPCR and its role in cancer progression has not been previously characterized. We found that GPR52 loss in breast cancer cells can lead to increased cell clustering, collective invasion, and EMT in vitro. These are features of increased cancer aggression. Our results reveal that GPR52 loss is a potential mechanism by which breast cancer progression may occur and support the investigation of GPR52 agonism as a therapeutic option for breast cancer. O_FIG O_LINKSMALLFIG WIDTH=169 HEIGHT=200 SRC="FIGDIR/small/604482v2_ufig1.gif" ALT="Figure 1"> View larger version (31K): org.highwire.dtl.DTLVardef@8b0f55org.highwire.dtl.DTLVardef@19bd139org.highwire.dtl.DTLVardef@c52117org.highwire.dtl.DTLVardef@1ffa2bd_HPS_FORMAT_FIGEXP M_FIG O_FLOATNOGraphical AbstractC_FLOATNO C_FIG

cancer biology↗