Biology subjects

Linial, M.

Publications and source records attributed to Linial, M..

At least 19 recordsLinked to original sources

bioRxiv2026

Spliceosomal miR-99b Regulates SPACA6-AS1 Pre-mRNA Levels and Promotes Malignant Phenotypes in Breast Cancer

MicroRNAs (miRNAs) are small non-coding RNAs that regulate gene expression primarily in the cytoplasm. However, emerging evidence highlights their additional roles in the nucleus. In particular, spliceosomal miRNAs have been implicated in novel regulatory functions, including the modulation of gene expression. Here, we investigate the nuclear role of spliceosomal miR-99b in breast cancer cells, focusing on its interaction with the long non-coding RNA (lncRNA) SPACA6-AS1. Using non-tumorigenic (MCF-10A) and breast cancer cell lines (MCF-7 and MDA-MB-231), we demonstrate that spliceosomal miR-99b expression increases with malignancy and correlates with elevated SPACA6-AS1 pre-mRNA levels. Notably, miR-99b exhibits full complementarity to the 5-prime splice junction of SPACA6-AS1, suggesting a direct role in splicing regulation. Functional assays reveal that inhibition of miR-99b reduces SPACA6-AS1 pre-mRNA levels, whereas its overexpression enhances pre-mRNA accumulation, indicating that miR-99b promotes the formation or stabilization of the unspliced transcript. Furthermore, increased miR-99b expression is associated with altered ratios of SPACA6 isoforms, supporting a broader role in RNA-level regulation of gene expression. Phenotypically, miR-99b enhances breast cancer cell migration and is required for efficient invasion, particularly in highly aggressive cancerous cells. Our findings uncover a novel nuclear function of miR-99b in modulating lncRNA splicing and gene expression. This spliceosomal miR-99b-SPACA6-AS1 axis represents a previously unrecognized regulatory pathway that contributes to breast cancer progression and may provide a potential target for diagnostic and therapeutic strategies.

molecular biology↗

bioRxiv2026

DJ-1/PARK7 Determines miRNA Network Plasticity During Genotoxic Stress

PARK7 (DJ-1) is a redox-sensitive stress-response protein that supports cellular adaptation, but its role in post-transcriptional responses to genotoxic stress remains unclear. We investigated whether DJ-1 abundance determines the miRNA response to X-ray-induced DNA damage. Integrated mRNA-seq and small RNA-seq across DJ-1 states in HEK293 cells revealed a striking divergence in miRNA plasticity. DJ-1 depletion by siRNA produced minimal miRNA remodeling, with only 34 (6.3%) miRNAs differentially expressed after irradiation. In contrast, elevated DJ-1 markedly increased miRNA plasticity: irradiation altered [~]37% of detectable miRNAs, accounting for [~]90% of miRNA reads, and extensively redistributed the miRNA pool. DJ-1 overexpression was also associated with remodeling of the miRNA regulatory machinery, particularly components involved in miRNA sorting and stability, suggesting feedback regulation of the miRNA pool. Comparison of precursor and mature species revealed substantial uncoupling between transcription and mature miRNA abundance, implicating regulation at the levels of processing, maturation, or stability. Radiation-responsive coding genes in DJ-1-overexpressing cells were relatively depleted of miRNA binding sites, supporting preferential regulation of upstream regulatory nodes rather than the bulk transcriptome. Together, these findings identify DJ-1 as a determinant of post-transcriptional signaling plasticity, enabling dynamic remodeling of the miRNA regulatory state in response to genotoxic stress. Graphical Abstract O_FIG O_LINKSMALLFIG WIDTH=200 HEIGHT=133 SRC="FIGDIR/small/744036v1_ufig1.gif" ALT="Figure 1000"> View larger version (38K): org.highwire.dtl.DTLVardef@18824aorg.highwire.dtl.DTLVardef@111d8bcorg.highwire.dtl.DTLVardef@ac33b4org.highwire.dtl.DTLVardef@17698d3_HPS_FORMAT_FIGEXP M_FIG C_FIG

bioinformatics↗

bioRxiv2026

MAGI: Mechanistic Consequences of Genetic Variants via Genomic Foundation Models

Clinical variant interpretation requires mechanism-aware evidence to guide diagnosis and clarify the biological consequences of mutations. However, existing computational predictors and genomic foundation models largely function as black boxes, providing pathogenicity labels with limited mechanistic insight or clinical actionability. Here, we present MAGI (Mechanistic Annotation of Genomic Impacts), a novel method that bridges this interpretability gap by unifying clinically relevant variant interpretation with mechanistic genomic analysis. MAGI pipeline leverages a genomic transformer model to quantify the effects of DNA variants across 3,623 functional tracks, encompassing regulatory features, multi-omics datasets, including tissue specificity and chromatin states, and 21 additional molecular annotations of genes and transcripts. These signals are integrated through a deterministic logic layer that maps single-nucleotide variants and indels to explicit molecular consequences. We benchmark MAGI-derived consequences against clinical rationales curated from ClinVar and observe strong concordance that scales with the magnitude of functional disruption. MAGI accurately recapitulates canonical pathogenic mechanisms, including start codon loss, splice site disruption, and regulatory element perturbation, consistent with ClinVar annotations. We further present case studies addressing conflicting or incomplete mechanistic interpretations, as well as variants requiring complex inference. Notably, MAGI is also applicable to non-human genomes and was evaluated on multispecies OMIA pathogenic variants. Collectively, MAGI establishes a generalizable framework that extends beyond clinical diagnostics to enable mechanistic discovery in functional genomics, generating mechanistically grounded, testable hypotheses for variants of uncertain significance (VUS) and variants with discordant clinical interpretations. In several cases, MAGI proposes alternative explanations that challenge existing annotations, providing transparent rationales and experimentally tractable predictions.

genetics↗

bioRxiv2025

OTRec: Deep learning recommender for prospective druggable disease-target associations

Identifying druggable disease--target associations remains a central challenge in translational medicine, limiting therapeutic discovery and repurposing. Here, we present OTRec, a deep learning--based recommender system that ranks such associations at scale and evaluates them in a temporal hold-out setting. Unlike approaches that rely on manually curated or aggregated evidence scores, OTRec employs a two-tower architecture to learn latent representations from 663,351 disease--target pairs. The model integrates heterogeneous inputs, including textual descriptions, ontology-derived features, and biological annotations such as tractability, Gene Ontology (GO) terms, and pathway information. We perform temporal validation by training on the 2022 Open Targets (OT) release and evaluating on clinical trial data from 2025. OTRec improves on the retrospective OT association score (ROC-AUC: 0.872 {+/-} 0.005 vs 0.559; PR-AUC: 0.288 {+/-} 0.009 vs. 0.08). In 5x5 target-disjoint cross-validation, OTRec reaches ROC-AUC 0.950 and PR-AUC 0.844) improving on the OT evidence score (ROC-AUC 0.91; PR-AUC 0.45). We rank the druggable genome across ~19,000 OT platform (OTP) diseases and release ~282,500 candidate associations above a 0.65 score threshold (in-distribution CV precision 0.92), covering 4,346 diseases including 2,322 orphan diseases, through an interactive prediction platform.

bioinformatics↗

bioRxiv2025

Signatures of Micropeptides Encoded by lncRNAs in Cancer Progression and Metastasis

Long non-coding RNAs (lncRNAs) are key regulators of gene expression, chromatin remodeling, and signaling. Recent estimates suggest that the human genome contains more than 35,000 lncRNA genes, with roughly 20% predicted to encode micropeptides (MPs) with unknown functions. In this study, we focused on the subset of lncRNAs with strong statistical evidence for MP-encoding potential, accounting for approximately 8% of the unfiltered MPs collection. Our analysis centered on 1,782 high-confidence lncRNA-MPs derived from 478 genes expressed across 17 cancer types from The Cancer Genome Atlas (TCGA). We show that lncRNA-MPs display distinct amino acid compositions and unique 4-mer patterns compared to the human coding proteome. A few genes (9) with exceptionally long transcripts are characterized by [≥]20 MPs each. Functional interference confirmed that most of the lncRNA-MPs are unstructured. Only a third of the genes display some phylogenetic conservation, and only 4 genes display canonical N-terminal signal peptides characteristic of secreted proteins. We focused on cancer progression-associated lncRNAs that show differential expression (z-score >|3|) across consecutive tumor stages and metastatic states (transitional lncRNAs, Tr-lncRNAs). A collection of 72 genes encoding 314 MPs (Tr-lncRNA-MPs) was detected, with 76% of the MPs being [≥]30 amino acids long. Prediction by AlphaFold 2.0 and homology modeling tools revealed dozens of MPs with well-defined secondary structures and recognizable 3D motifs. Among the longer Tr-lncRNA-MPs (>60 amino acids), we confirmed the presence of ubiquitin-like, RNase H-related, and other conserved foldable motifs. Known cancer lncRNAs containing high-confidence MPs (XIST, UCA1, HOXA11-AS, LINC01234, and HAND-AS1) overlap with 50 pan-cancer lncRNAs associated with tumor stage or metastasis transitions. Together, these findings demonstrate that integrating sequence motifs (e.g., signal peptides, k-mers) with structural foldability offers a multifaceted view of lncRNA-MPs in cancer. We argue that the capacity to produce MPs may reinforce the oncogenic impact dominated by the lncRNA entity. We propose that Tr-lncRNA-MPs represent a promising new class of biomarkers and therapeutic targets in oncology. Key pointsO_LI478 lncRNA genes with strong evidence for micropeptide (MPs) production generated 1,782 distinct lncRNA-MPs. C_LIO_LI72 lncRNAs and 314 MPs are associated with transitional lncRNAs from 17 cancer types and stages of tumor progression and metastasis. C_LIO_LISequence and structural analyses reveal many MPs with reliable 3D folding potential. C_LIO_LIDozens of previously overlooked MPs may serve as novel biomarkers and therapeutic targets in cancer. C_LI

cancer biology↗

bioRxiv2025

Deciphering the mechanistic basis for the pathological effect of the Gαo E246K mutation in neurodevelopmental disorder

Mutations in the GNAO1 gene, which encodes for Go, a major neuronal G protein, are associated with neurodevelopmental disorders, epilepsy, and movement disorders. We identified and characterized a spontaneous heterozygous GNAO1 E246K mutation in an Israeli female infant with complex developmental delays and substantial motor difficulties. This mutation has been reported in other cases as a prevalent pathogenic mutation in patients with motor dysfunction and a broad range of neurological outcomes. To investigate the molecular and functional consequences of the Go E246K mutation, we employed structural modeling and analysis, biochemical assays, mass spectrometry-based proteomics, and cellular functional assays. We show that this mutation does not affect nucleotide binding, nor basal or RGS- accelerated GTP hydrolysis. Despite the E246 position located within a predicted effector binding region, proteomics analysis did not identify any new cellular partners. Instead, we demonstrate that the E246K mutation disrupts the Go regulatory GTPase cycle by directly impairing G{beta}{gamma} dissociation. This impairment overrides the presence of wild-type Go, explaining the dominant effect of the severe neurogenetic phenotype in the heterozygous background. These findings establish a new molecular mechanism for a GNAO1 mutation with dominant-negative effects on the GTPase regulatory cycle. The insights gained from studying this mechanism of action provide a basis for developing specific and personalized therapeutic strategies based on the outcome of a missense mutation in GNAO1.

neuroscience↗

bioRxiv2025

Transitional lncRNA Signatures Reveal Distinct Stages of Cancer Progression and Metastasis

Long non-coding RNAs (lncRNAs) are emerging as key regulators in cancer, influencing gene expression, chromatin remodeling, and signaling. Evidence from The Cancer Genome Atlas (TCGA) and other datasets supports their role in tumor progression. Although the human genome harbors thousands lncRNA genes, only a small subset has been validated in cancer. In this study, we used the LncBook catalog ([~]95,000 lncRNAs) to identify [~]12,500 lncRNAs with expression evidence across major TCGA cancer types. These were stratified by clinical annotations, including cancer stage (I-IV) and metastatic state (M0/M1). Using significant differential expression (z-score >|3|) for consecutive transitions, we identified a set of influential transitional lncRNAs (Tr-lncRNAs) that signify cancer transitions. Analyzing seven transitions revealed that over 70% of Tr-lncRNAs were cancer-type specific, while only 2-4% were shared across 17 major cancers. Each cancer type had 30-80 Tr-lncRNAs, with more than half uniquely expressed in one type. Most Tr-lncRNAs were previously uncharacterized. A pan-cancer analysis revealed 14 shared Tr-lncRNAs, including known ones such as XIST and H19. Our findings highlight distinct lncRNA expression patterns during cancer progression and provide new insights into cis-regulatory antisense mechanisms. We discuss the potential of Tr-lncRNAs as diagnostic biomarkers and therapeutic targets in cancer.

cancer biology↗

bioRxiv2025

Temporal shifts in microRNAs signify the inflammatory state of primary murine microglial cells

The primary function of microglia is to maintain brain homeostasis. However, in several neurodegenerative diseases, such as Alzheimers disease, the pathophysiological hallmarks that drive disease progression involve neurotoxicity and alterations in neuroinflammation. In this study, we exposed murine neonatal primary microglial cultures to external signals that mimic the in vivo stimuli caused by pathogens, injury, or toxic agents. In the presence of benzoyl ATP (bzATP) and lipopolysaccharide (LPS), we observed a coordinated increase in the expression of interleukins and chemokines. We focused on the dynamics of the differentially expressed microRNAs (miRNAs) that are statistically significant (DEMs) and tracked their post-activation dynamics. Monitoring miRNAs 3 and 8 hours (h) post-activation revealed robust changes in 33 and 57 DEMs, most of which were upregulated. The DEMs exhibiting the strongest temporal regulation included miR-155, miR-132, miR-3473e, miR-222, and miR-146b. Additionally, a strong downregulation of miR-3963 was attributed to the exposure to bzATP. Through their regulation of TNF and NF{kappa}B signaling pathways, the identified DEMs reflect the cellular response to inflammatory signals. We incubated the activated cells with ladostigil, a neuroprotective compound that has been shown to reduce oxidative stress, inflammation, and cognitive decline. While there was no significant effect 3 h post-activation, at 8 h, a few miRNAs implicated in inflammation suppression such as miR-27a, miR-27b, and miR-23b were upregulated in a ladostigil-dependent manner. We conclude that the miRNA expression profile provides a sensitive indicator of the regulatory mechanisms underlying inflammation-related responses in microglia. We propose that primary microglia subjected to controlled activation paradigms can serve as a robust model for inflammatory states in the brain to monitor the aging brain along the progression of neurodegenerative diseases.

cell biology↗

bioRxiv2025

Patient-Specific Pharmacogenomics unveils xCT Regulation Pathways in Colon Cancer

Colorectal cancer (CRC) represents the third-leading cause of cancer-related deaths. Knowledge covering diverse cellular and molecular data from individual patients has become valuable for diagnosis, prognosis, and treatment selection. Here, we present an in-depth comparative mRNA-seq and microRNA-seq analysis of tissue samples from 32 CRC, pairing tumors with adjacent healthy tissues. The differential expression gene (DEG) analysis revealed an interconnection between nutrients, metabolic programs, and cell cycle pathways. We focused on the impact of overexpressed SLC7A11 (xCT) and SLC3A2 genes which compose the cystine/glutamate transporter (Xc-) system. To assess the oncogenic potency of the Xc-system in a cellular setting, we applied a knowledge-based approach for analyzing gene perturbations from CRISPR screens across various cell types as well as using a variety of functional assays in five primary patient-derived organoid cell models to functionally verify our hypothesis. We identified a previously undescribed cell surface protein signature predicting chemotherapy resistance and further highlighted the causality and potential of pharmacological blockage of ferroptosis as promising avenue for cancer therapy. Biological processes such as redox homeostasis, ion/amino acid transporters and de novo nucleotide synthesis were associated with these co-dependent genes in patient specimens. This study highlighted a number of overlooked genes as potential clinical targets for CRC and promotes stem cell-based, patient-individual in vitro model systems as a versatile partner platform to functionally validate in silico predictions, with focus on SLC7A11 and its associated genes in tumorigenesis.

cancer biology↗

bioRxiv2025

Manipulating PARK7/DJ-1 Levels by Genotoxic Stress Alters Noncoding RNAs and Cellular Homeostasis

DJ-1/PARK7 is a multifunctional protein that plays a vital role in sensing oxidative stress and maintaining redox homeostasis. As an oncogene, DJ-1 influences p53-mediated stress responses and contributes to cancer progression. This study investigates the impact of X-ray-induced DNA breaks on cellular responses post-irradiation under varying DJ-1 expression levels. Using siRNA knockdown and overexpression to modulate DJ-1 levels, transcriptional changes were analyzed through RNA-seq. Naive cells exhibited only a moderate transcriptional response to X-ray exposure, including suppression of the cell cycle and activation of stress pathways. Overexpression of DJ-1 led to pronounced gene expression suppression, particularly suppression of most ribosomal and mitochondrial genes. Conversely, DJ-1 knockdown caused extensive, non-specific transcriptional changes, indicating disrupted cellular homeostasis. Notably, around 25% of non-coding RNAs were differentially expressed under these conditions. DJ-1 overexpression suppressed lncRNA of the host genes for snoRNAs, thus altering the miRNA sponging capacity after X-ray exposure. These findings underscore DJ-1s critical role in modulating cellular responses to genotoxic stress, reshaping transcriptional landscapes, and regulating non-coding RNA profiles. DJ-1 emerges as essential for maintaining genomic stability.

molecular biology↗

bioRxiv2024

Knockdown of DJ-1 Resulted in a Coordinated Activation of the Innate Immune Antiviral Response in HEK293 Cell Line

PARK7, also known as DJ-1, plays a critical role in protecting cells by functioning as a sensitive oxidation sensor and modulator of antioxidants. DJ-1 acts to maintain mitochondrial function and regulate transcription in response to different stressors. In this study, we show that cell lines vary by their antioxidation potential at basal condition. The transcriptome of HEK293 cells was tested following knockdown (KD) of DJ-1 using siRNAs which reduced the DJ-1 transcripts to only 12% of the original level. We compared the expression levels of 14k protein coding transcripts, and 4.2k non-coding RNAs relative to control cells treated with non-specific siRNAs. Among the coding genes, [~]200 upregulated differentially expressed genes (DEGs) signified a coordinated antiviral innate immune response. Most genes were associated with regulation of type 1 interferons (IFN) and induction of inflammatory cytokines. About a quarter of these genes were also induced in cells treated by non-specific siRNAs that were used as a negative control. Beyond the antiviral response, 114 genes were specific to KD of DJ-1 with enrichment in RNA metabolism and mitochondrial functions. A smaller set of downregulated genes (58 genes) were associated with dysregulation in membrane structure, cell viability, and mitophagy. We propose that KD of DJ-1 diminish its protective potency against oxidative stress, rendering the cells labile and responsive to dsRNA signal by activation of a large number of genes, many of which drive apoptosis, cell death, and inflammatory signatures. The KD of DJ-1 highlights its crucial role in regulating genes associated with antiviral responses, RNA metabolism, and mitochondrial functions, apparently through alteration in STAT activity and downstream signaling. Given that DJ-1 is highly expressed in metastatic cancers, targeting DJ-1 could be a promising therapeutic strategy where manipulation of DJ-1 level may reduce cancer cell viability and enhance the efficacy of cancer treatments.

bioinformatics↗

bioRxiv2024

Transcription network of SLC7A11 (xCT) in colon cancer provides clinical targets for metabolic regulation and cell proliferation

Colorectal cancer (CRC) represents the third leading cause of cancer-related deaths. Knowledge covering diverse cellular and molecular data from individual patients has become valuable for diagnosis, prognosis, and treatment selection. Here, we present in-depth comparative RNA-seq analysis of 32 CRC patients pairing tumor and healthy tissues (total of 73 samples). Strict thresholds for differential expression genes (DEG) analysis revealed an interconnection between nutrients, metabolic program, and cell cycle pathways. Among the upregulated DEGs, we focused on the Xc- system, composed of the proteins from SLC7A11 (xCT) and SLC3A2 genes, along with several interacting genes. To assess the oncogenic potency of the Xc- system in a cellular setting, we applied a knowledge-based approach, analyzing gene perturbations from CRISPR screens. The study focused on a set of 27 co-dependent genes that were strongly correlated with the fitness of SLC7A11 and SLC3A2 across many cell types. Alterations in these genes in 13 large-scale studies (e.g., by mutations and copy number variation) were found to enhance overall survival and progression-free survival in CRC patients. In agreement, the overexpression of these genes in cancer cells drives cancer progression by allowing effective management of the redox level, induction of stress response mechanisms, and most notably, enhanced activity of ion/amino acid transporters, and enzymes acting in de novo nucleotide synthesis. We also highlight the positive correlation between the Xc- system gene expression level, patient responsiveness to different chemotherapy treatments, and immune cell infiltration (e.g., myeloid-derived suppressor cells) in CRC tumors as a measure for their immunosuppressive activity. This study illustrates that knowledge-based interpretation by synthesizing multiple layers of data leads to functional and mechanistic insights into the role of SLC7A11 and its associated genes in CRC tumorigenesis and therapeutics.

cancer biology↗

bioRxiv2024

Protein Language Models Expose Viral Mimicryand Immune Escape

MotivationViruses elude the immune system through molecular mimicry, adopting biophysical characteristics of their host. We adapt protein language models (PLMs) to differentiate between human and viral proteins. Understanding where the immune system and our models make mistakes could reveal viral immune escape mechanisms. ResultsWe applied pretrained deep-learning PLMs to predict viral from human proteins. Our predictors show state-of-the-art results with AUC of 99.7%. We use interpretable error analysis models to characterize viral escapers. Altogether, mistakes account for 3.9% of the sequences with viral proteins being disproportionally misclassified. Analysis of external variables, including taxonomy and functional annotations, indicated that errors typically involve proteins with low immunogenic potential, viruses specific to human hosts, and those using reverse-transcriptase enzymes for their replication. Viral families causing chronic infections and immune evasion are further enriched and their protein mimicry potential is discussed. We provide insights into viral adaptation strategies and highlight the combined potential of PLMs and explainable AI in uncovering mechanisms of viral immune escape, contributing to vaccine design and antiviral research. Availability and implementationData and results available in https://github.com/ddofer/ProteinHumVir. Contactmichall@cc.huji.ac.il

bioinformatics↗

bioRxiv2023

The Interactome of DUX4 Reveals Multiple Activation Pathways

The ectopic expression of the DUX4 protein in muscle cells is the underlying cause of Facioscapulohumeral Muscular Dystrophy (FSHD). DUX4 is a potent transcription factor that activates a large number of genes in a dysregulated manner, but the direct protein interactions involved in this activation are only partially known. Here, we tailored an affinity purification and mass spectrometry (AP-MS) analysis to the unique features and functions of DUX4, to provide a more complete view of its interactome. We also obtained and verified all-atom models for two of the major interactions by employing cross-linking and mass spectrometry (CL-MS), computational modeling, and guided mutation studies. We find that DUX4 interacts strongly with two homologous transcription activators, PTOV1 and MED25, in addition to the previously characterized CBP/p300. The interaction with the PTOV1/MED25 domain involves the wrapping of the last thirty residues of the DUX4 activation region around the domain in a very extensive interface. Hence, DUX4 has the capacity to both open the chromatin and directly recruit the Mediator complex. DUX4 also binds to all members of the RFPL4 family, which are among the strongest genes it activates. These interactions are mediated through a hitherto unrecognized motif in the DUX4 disordered linker region. This feedback mechanism suggests that DUX4 may be inhibited by its own activation products, and explains its typical pulsed expression profile. We also found SIX1 and the AP-2 complex as strong DUX4 C-terminal interactors. A separate analysis of interactions involving the N-terminal of DUX4 revealed enrichment of proteins that are involved in DNA repair following double-strand breaks. Overall, these findings reveal new activation pathways for DUX4, which may be modulated in future strategies to control its toxicity. This study also showcases the synergy between CL-MS and deep-learning based modeling for the structural elucidation of challenging protein-protein interactions.

biochemistry↗

bioRxiv2023

Detecting Anomalous Proteins Using Deep Representations

Many advances in biomedicine can be attributed to identifying unusual proteins and genes. Many of these proteins unique properties were discovered by manual inspection, which is becoming infeasible at the scale of modern protein datasets. Here, we propose to tackle this challenge using anomaly detection methods that automatically identify unexpected properties. We adopt a state-of-the-art anomaly detection paradigm from computer vision, to highlight unusual proteins. We generate meaningful representations without labeled inputs, using pretrained deep neural network models. We apply these protein language models (pLM) to detect anomalies in function, phylogenetic families, and segmentation tasks. We compute protein anomaly scores to highlight human prion-like proteins, distinguish viral proteins from their host proteome, and mark non-classical ion/metal binding proteins and enzymes. Other tasks concern segmentation of protein sequences into folded and unstructured regions. We provide candidates for rare functionality (e.g., prion proteins). Additionally, we show the anomaly score is useful in 3D folding-related segmentation. Our novel method shows improved performance over strong baselines and has objectively high performance across a variety of tasks. We conclude that the combination of pLM and anomaly detection techniques is a valid method for discovering a range of global and local protein characteristics.

bioinformatics↗

bioRxiv2023

Selecting Covariates for Genome-Wide Association Studies

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

genetics↗

bioRxiv2022

Inferring microRNA regulation: A proteome perspective

Post-transcriptional regulation in multicellular organisms is mediated by microRNAs. However, the mechanisms that determine if a gene is regulated by miRNAs are poorly understood. Previous works focused mostly on miRNA seed matches and other features of the 3-UTR of transcripts. These common approaches relied on knowledge of the miRNA families, and computational approaches still yield poor, inconsistent results, with many false positives. In this work, we present a different paradigm for predicting miRNA-regulated genes based on proteins. In a novel, automated machine learning framework, we use sequence as well as diverse functional annotations to train models on multiple organisms using experimentally validated data. We present insights from tens of millions of features extracted and ranked from different modalities. We show high predictive performance per organism and in generalization across species. We provide a list of novel predictions for Danio rerio (zebrafish) and Arabidopsis thaliana (mouse-ear cress). We compare genomic models, and observe that our protein model outperforms genomics, while a unified model improves on both. While most membranous and disease related proteins are regulated by miRNAs, we observe the G-protein coupled receptor (GPCR) family is an exception, being mostly unregulated by miRNAs, and we raise possible explanations for this. We further show that the evolutionary conservation among duplicated genes does not imply a coherence in miRNA regulation. We conclude that duplicated genes diverge in their tendency to be miRNA regulated. However, protein function is informative across species in predicting post-transcriptional miRNA regulation in living cells.

bioinformatics↗

bioRxiv2022

Alteration in long noncoding RNAs in response to oxidative stress and ladostigil in SH-SY5Y cells

Microglia activation causes neuroinflammation, which is a hallmark of neurodegenerative disorders, brain injury, and aging. Ladostigil, a bifunctional reagent with antioxidant and anti-inflammatory properties, reduced microglial activation and enhanced brain functioning in elderly rats. In this study, we studied SH-SY5Y, a human neuroblastoma cell line, and tested viability in the presence of hydrogen peroxide and Sin1 (3-morpholinosydnonimine), which generates reactive oxygen and nitrogen species (ROS/RNS). Both stressors caused significant apoptosis and necrotic cell death that was attenuated by ladostigil. Our results from RNA-seq experiments show that long non-coding RNAs (lncRNAs) account for 30% of all transcripts in SH-SY5Y cells treated with Sin1 for 24 hours. Altogether, we identify 94 differently expressed lncRNAs in the presence of Sin1, including MALAT1, a highly expressed lncRNA with anti-inflammatory and anti-apoptotic functions. Additional activities of Sin-1 upregulated lncRNAs include redox homeostasis (e.g., MIAT, GABPB1-AS1), energy metabolism (HAND2-AS1), and neurodegeneration (e.g., MIAT, GABPB1-AS1, NEAT1). Four lncRNAs implicated as enhancers were significantly upregulated in cells exposed to Sin1 and ladostigil. Finally, we show that H2O2 and Sin1 increased the expression of DJ-1, a redox sensor and modulator of Nrf2 (nuclear factor erythroid 2- related factor 2). Nrf2 (NFE2L2 gene) is a major transcription factor regulating antioxidant genes. In the presence of ladostigil, DJ-1 expression is restored to its baseline. The mechanisms governing SH-SY5Y cell survival and homeostasis are highlighted by the beneficial role of ladostigil in the crosstalk involving Nrf2, antioxidant transcription factor DJ-1, and lncRNAs. Stress-dependent induction of lncRNAs represents an underappreciated regulatory level that contributes to cellular homeostasis and the capacity of SH-SY5Y to cope with oxidative stress.

cell biology↗