Biology subjects

Parthasarathy, S.

Publications and source records attributed to Parthasarathy, S..

At least 19 recordsLinked to original sources

bioRxiv2026

Multi-Trait Meta-QTL Analysis Reveals Genomic Hotspot Classes for Strategic Maize Improvement

BackgroundDecades of maize (Zea mays L.) quantitative trait locus (QTL) mapping have produced fragmented results across hundreds of independent studies, characterized by broad confidence intervals, population-specific effects, and a predominantly single-trait scope. Comprehensive multi-trait integration remains limited in the public domain, yet it could improve our understanding of trait relationships for strategic maize breeding. We integrated 2,701 QTLs published over 30 years across five trait categories (grain yield and components (GYC); plant development and architecture (PDA); plant physiology and stress adaptation (PHS); grain quality and nutritional composition (GQN); and disease and pest resistance (DPR)) to identify functionally classified genomic hotspots and prioritize candidate genes for multi-trait breeding. ResultsBioMercator V4.2.3 consolidated 2,518 projectable QTLs into 187 high-confidence meta-QTLs (MQTLs), achieving a 4.0-fold (75%) reduction in mean genetic-map confidence interval width (47.6 to 11.8 cM); 128 of 187 MQTLs (68.4%) showed genome-wide association study (GWAS) co-localization, and 119 (63.6%) met the strict dual-platform criterion. Twenty-three genomic hotspots harbored 131 of 187 MQTLs (70.1%) and were classified by an ordered, data-driven rule into three categories: ten multi-trait hubs enabling simultaneous improvement through pleiotropic or tightly linked genes; seven pathway-focused clusters dominated by a single trait pathway, exemplified by the chromosome 8 kernel-yield cluster; and six major-effect loci, each carrying a member meta-QTL of large effect (phenotypic variance explained (PVE) of at least 28%), including the Rp1 common-rust locus (35% member-MQTL PVE) and the chromosome 9 Wx1/THP9 starch-and-protein locus (up to 52% member-MQTL PVE). Environmental classification distinguished MQTLs predominantly supported by optimal-condition QTLs (42%) from those supported by stress-condition QTLs (28%), the latter showing [~]1.7-fold greater mean contributing-QTL phenotypic variance (12.8% vs. 7.4%), consistent with conditional effect amplification under stress; within-population comparisons under contrasting water regimes showed a larger, [~]3.5-fold amplification. Network-based candidate gene prioritization with cross-cereal ortholog analysis identified priority candidates with confirmed orthologs in one or more comparison cereals (rice, sorghum, wheat, or barley), with conservation highest for developmental and metabolic genes and lowest for pathogen-resistance genes, identifying priority targets for functional genomics investment. ConclusionsThis functionally classified and environmentally characterized meta-QTL framework provides breeders with a structured resource for multi-trait hotspot selection, environment-appropriate allele deployment, and functional genomics prioritization, with broader applicability as a transferable template for other crops confronting fragmented QTL literature and complex multi-trait breeding objectives.

plant biology↗

bioRxiv2026

Simultaneous broad protection against Ebola Sudan, Marburg and Lassa viruses conferred by a DNA primed MVA-vectored multivalent vaccine

Sub-Saharan Africa continues to experience recurrent outbreaks of zoonotic viral diseases that spill over unpredictably from animal reservoirs into human populations. In many regions, mpox co-circulates with viral hemorrhagic fevers (VHFs) caused by Ebola Sudan virus (SUDV), Marburg virus (MARV), and Lassa fever virus (LASV). Overlapping clinical syndromes that these VHF cause challenge surveillance, diagnostics and timely deployment of effective countermeasures. A single vaccine capable of protecting against these biologically and genetically distinct pathogens would markedly reduce the cost and complexity of outbreak response, lessen dependence on emergency international aid, and strengthen long-term health system resilience. Here, we report on the development of an MVA-based mpox vaccine engineered to express computationally designed, broad-coverage antigens targeting SUDV, MARV and LASV. In preclinical challenge studies, this multivalent vaccine elicited robust immune responses and conferred significant protection against lethal infection from all three pathogens in parallel challenge experiments. These findings establish preclinical proof-of-concept for a single, broadly protective VHF vaccine and support its clinical development for deployment across diverse settings in Sub-Saharan Africa. SignificanceOutbreak control in Sub-Saharan Africa is challenged by the co-circulation of different high consequence human infections such as mpox and diverse viral hemorrhagic fevers (VHF) such as SUDV, MARV, and LASV pathogens. These VHFs have overlapping early clinical syndromes, complicating triage and delaying effective targeted interventions. We developed a single MVA-based vaccine encoding computationally designed, conserved antigens from all three VHFs encoded within the MVA vector analogous to the licensed mpox vaccine. In simultaneous challenge models, this multivalent vaccine elicited robust humoral and cellular responses and conferred significant protection against lethal infection by each hemorrhagic fever pathogen. This work provides preclinical proof-of-concept for a unified, broadly protective countermeasure compatible with existing MVA-mpox vaccine manufacturing and deployment experience. By reducing dependence on rapid differential diagnostics and streamlining logistics relative to maintaining multiple pathogen-specific vaccine stockpiles, this approach can lower costs, accelerate response, and increase equity of access during syndromic outbreaks. The platforms engineered antigen breadth and human safety profile of MVA together support a pragmatic translational pathway toward clinical evaluation and regional readiness for zoonotic spillover events that are intensifying with human and environmental changes.

immunology↗

bioRxiv2025

The neural mechanisms supporting the rise and fall of maternal aggression

To protect the helpless young, lactating females dramatically increase aggression towards intruders, known as maternal aggression. However, attack is costly and risky. When pups no longer exist, maternal aggression rapidly declines. Our study reveals the critical role of the pathway from posterior amygdala cells expressing estrogen receptor alpha (PAEsr1) to the ventrolateral part of ventromedial hypothalamus cells expressing neuropeptide Y receptor Y2 (VMHvlNpy2r) in the rise and fall of maternal aggression. Functional manipulations and recordings demonstrate that VMHvl-projecting PAEsr1 (PAEsr1[->]VMHvl) cells are naturally active and required for maternal aggression. During lactation, PA-VMHvlNpy2r connection strengthens, and VMHvlNpy2r excitability increases to facilitate attack. Furthermore, PAEsr1 expresses abundant oxytocin receptors, enabling oxytocin to boost PAEsr1 cell output. After pup separation, the oxytocin level drops, causing decreased maternal aggression, which can be restored by optogenetically increasing the oxytocin level. Thus, diverse forms of plasticity occur at the PAEsr1-VMHvlNpy2r circuit to support need-based maternal aggression.

neuroscience↗

bioRxiv2025

PURE: Policy-guided Unbiased REpresentations for structure-constrained molecular generation

Structure-constrained molecular generation (SCMG) generates novel molecules that are structurally similar to a given molecule and have optimized properties. Deep learning solutions for SCMG are limited in that they are pre-disposed towards existing knowledge, and they suffer from a natural impedance mismatch problem due to the discrete nature of molecules, while deep learning methods for SCMG often operate in continuous space. Moreover, many task-specific evaluation metrics used during training often bias the model towards a particular metric -"metric-leakage". To overcome these shortcomings, we propose Policy-guided Unbiased REpresentations (PURE) for SCMG that learn within a framework simulating molecular transformations for drug synthesis. PURE combines self-supervised learning with a policy-based reinforcement-learning (RL) framework, thereby avoiding the need for external molecular metrics while learning high-quality representations that incorporate an inherent notion of similarity specific to the given task. Along with a semi-supervised training design, PURE utilizes template-based molecular simulations to better explore and navigate the discrete molecular search space. Despite the lack of metric biases, PURE achieves competitive or superior performance than state-of-the-art methods on multiple benchmarks. Our study emphasizes the importance of reevaluating current approaches for SCMG and developing strategies that naturally align with the problem. Finally, we illustrate how our methodology can be applied to combat drug resistance, by identifying sorafenib-like compounds as a case study.

biochemistry↗

bioRxiv2025

Covering All Bases: A Computational Method to Design Broad-spectrum T-cell-inducing Vaccines Applied to Betacoronaviruses

Antigenically diverse pathogens, such as coronaviruses, pose substantial threats to global health. This highlights the need for effective broad-spectrum vaccines that elicit robust immune responses in a large proportion of the human population against a wide array of pathogen variants. Here, we introduce Spectravax, an AI-enabled computational method to design broad-spectrum vaccines that account for genetic diversity in both the host and pathogen populations. Using Spectravax, we designed a nucleocapsid (N) antigen to elicit cross-reactive immune responses to viruses from the Sarbecovirus and Merbecovirus subgenera of Betacoronaviruses. In silico analyses demonstrated superior predicted host and pathogen coverage for Spectravax compared to wild-type sequences and existing computational designs. Experimental validation in mice supported these predictions: Spectravax N elicited robust immune responses to SARS-CoV, SARS-CoV-2, and MERS-CoV--the three coronaviruses responsible for major outbreaks in humans since 2002--while wild-type and existing computational designs elicited limited responses. Furthermore, we were able to identify the MERS-CoV N epitopes responsible for Spectravaxs cross-reactivity. Thus, we advance the rational design of broad-spectrum vaccines for pandemic preparedness.

bioinformatics↗

bioRxiv2025

Phenotypic analysis of 11,125 trio exomes in neurodevelopmental disorders

Genomic sequencing is widely used to identify causative genetic changes in neurodevelopmental disorders, such as autism, intellectual disability, and epilepsy. Most neurodevelopmental disorders also present with diverse clinical features, and delineating the interaction between causative genetic changes and phenotypic features is a key prerequisite for developing personalized therapies. However, assessing clinical features at a scale that parallels genomic sequencing remains challenging. Here, we standardize phenotypic information across 11,125 patient-parent trios with exome sequencing data using biomedical ontologies, analyzing 674,767 phenotypic terms. We find that individuals with de novo variants in 69 out of 261 neurodevelopmental genes exhibit statistically significant clinical similarities with distinct phenotypic fingerprints. We also observe that phenotypic relatedness follows a gradient, spanning from highly similar to dissimilar phenotypes, with intra-gene similarities suggesting clinically distinct subgroups for seven neurodevelopmental genes. For most genetic etiologies, only a small subset of highly phenotypically similar individuals carried de novo variants in the same gene, highlighting the heterogeneous and complex clinical landscape of neurodevelopmental disorders. Our study provides a large-scale overview of the dynamic relationship between genotypes and phenotypes in neurodevelopmental disorders, underscoring how the inherent complexity of these conditions can be deciphered through approaches that integrate genomic and phenotypic data.

genetics↗

bioRxiv2025

IFN-γ signaling is required for the efficient replication of murine hepatitis virus (MHV) strain JHM in the brains of infected mice.

Neurotropic viruses are a major public health concern as they can cause encephalitis and other severe brain diseases. Many of these viruses, including flaviviruses, herpesviruses, rhabdoviruses and alphaviruses enter the brain through the olfactory neuroepithelium (ONE) in the olfactory bulbs (OB). Due to the low percentage of encephalitis that occurs following these infections, its thought that the OBs have specialized innate immune responses to eliminate viruses. Murine hepatitis virus strain JHM (JHMV) is a model coronavirus that causes severe encephalitis in mice and can access the brain through olfactory sensory neurons. Weve shown that a JHMV Mac1-mutant virus, N1347A, has decreased replication and disease in the brains of mice. Here we further show that this virus replicates poorly in the OB. However, it is unknown which innate immune factors restrict N1347A replication in the OB. RNA seq analysis of infected olfactory bulbs showed that IFN{gamma} was upregulated in the OB while IFN-{beta} was barely detectable at 5 days post-infection. To determine if IFN-{gamma} restricts JHMV N1347A replication, we utilized IFN-{gamma} and IFN-{gamma} receptor (IFN-{gamma}R) knockout (KO) mice. Surprisingly we found that JHMV WT and N1347A replicated very poorly in the OB and whole brains of both IFN-{gamma} and IFN-{gamma}R KO mice following intranasal infection, though survival and weight loss were unaltered. Furthermore, we determined that microglia were the primary cells producing IFN-{gamma} during the early stages of this infection. We conclude that IFN-{gamma} is required for the efficient replication of JHMV in the brains of infected mice.

microbiology↗

bioRxiv2024

Optimizing clinical interpretability of functional evidence in epilepsy-related ion channel variants

Variants in genes encoding the voltage-gated ion channels are among the most common monogenic causes of epilepsy and neurodevelopmental disorders. Functional effects of a variant are increasingly important for diagnosis and therapeutic decisions. To incorporate knowledge regarding functional consequences in formal clinical variant interpretation, we developed an approach for evaluating multiple functional measurements within the Bayesian framework of the modified ACMG/AMP guidelines. We analyzed 216 functional assessments of 191 variants in SCN1A (n=74), SCN2A (n=66), SCN3A (n=18), and SCN8A (n=33). Of 20 commonly measured biophysical parameters, the most frequent drivers of overall functional consequence were persistent current (f=0.54), voltage dependence of activation (f=0.51), and voltage dependence of fast inactivation (f=0.40) for gain-of-function and peak current (f=0.87) for loss-of-function. By comparing measurements of 23 benign variants, we determined thresholds by which published data on these four parameters confer Strong evidence of variant pathogenicity (likelihood ratio > 18.7) under the ACMG/AMP rubric. Similarly, we delineated evidence weights for the most common epilepsy-related potassium channel gene, KCNQ2, through reports of 80 pathogenic and 24 benign variants, accounting for heterozygous and homozygous experimental conditions. We collected the resulting categorization of functional data into FENICS, a biomedical ontology of 152 standardized terms for coherent annotation of electrophysiological results. Across 271 variants in SCN1A/2A/3A/8A and KCNQ2, 1,731 annotations are available in ClinVar, facilitating use of this evidence in variant classification. In summary, we introduce and apply an ACMG/AMP-calibrated framework for electrophysiological studies in epilepsy-related channelopathies to delineate the impact of functional evidence on clinical variant interpretation.

genetics↗

bioRxiv2024

PARP14 is pro- and anti-viral host factor that promotes IFN production and affects the replication of multiple viruses

PARP14 is a 203 kDa multi-domain protein that is primarily known as an ADP-ribosyltransferase, and is involved in a variety of cellular functions including DNA damage, microglial activation, inflammation, and cancer progression. In addition, PARP14 is upregulated by interferon (IFN), indicating a role in the antiviral response. Furthermore, PARP14 has evolved under positive selection, again indicating that it is involved in host-pathogen conflict. We found that PARP14 is required for increased IFN-I production in response to coronavirus infection lacking ADP-ribosylhydrolase (ARH) activity and poly(I:C), however, whether it has direct antiviral function remains unclear. Here we demonstrate that the catalytic activity of PARP14 enhances IFN-{beta} and IFN-{lambda} responses and restricts ARH-deficient murine hepatitis virus (MHV) and severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) replication. To determine if PARP14s antiviral functions extended beyond CoVs, we tested the ability of herpes simplex virus 1 (HSV-1), a DNA virus, vesicular stomatitis virus (VSV), a negative-sense RNA virus, and lymphocytic choriomeningitis virus (LCMV), an ambisense RNA virus, to infect A549 PARP14 knockout (KO) cells. While LCMV infection was unaffected, HSV-1 replication was increased in PARP14 KO cells and VSV replication was decreased. These results indicate that PARP14 restricts HSV-1 replication but enhances the replication of VSV. A PARP14 active site inhibitor had no impact on HSV-1 or VSV replication, indicating that its effect on these viruses was independent of its catalytic activity. These data demonstrate that PARP14 promotes IFN production and has both proviral and antiviral functions targeting multiple viruses. IMPORTANCEThe antiviral response is largely regulated by post-translation modifications (PTM), including ADP-ribosylation. PARP14 is an ADP-ribosyltransferase that is upregulated by interferon and is under positive selection, indicating that it is involved in host-pathogen conflict. However, no anti-viral function has been described for PARP14. Here, we found that PARP14 represses both coronavirus and HSV-1 replication, demonstrating that PARP14 has antiviral functions. Surprisingly, we also found that PARP14 also has pro-viral functions, as it was critical for the efficient replication of VSV. These data indicate that PARP14 has both proviral and antiviral functions. Defining the mechanisms used by PARP14 to both repress and promote virus replication will provide new insights into how PARPs regulate virus infection.

microbiology↗

bioRxiv2023

PARP12 is required to repress the replication of a Mac1 mutant coronavirus in a cell and tissue specific manner

ADP-ribosyltransferases (ARTs) mediate the transfer of ADP-ribose from NAD+ to protein or nucleic acid substrates. This modification can be removed by several different types of proteins, including macrodomains. Several ARTs, also known as PARPs, are stimulated by interferon, indicating ADP-ribosylation is an important aspect of the innate immune response. All coronaviruses (CoVs) encode for a highly conserved macrodomain (Mac1) that is critical for CoVs to replicate and cause disease, indicating that ADP-ribosylation can effectively control coronavirus infection. Our siRNA screen indicated that PARP12 might inhibit the replication of a MHV Mac1 mutant virus in bone-marrow derived macrophages (BMDMs). To conclusively demonstrate that PARP12 is a key mediator of the antiviral response to CoVs both in cell culture and in vivo, we produced PARP12-/- mice and tested the ability of MHV A59 (hepatotropic/neurotropic) and JHM (neurotropic) Mac1 mutant viruses to replicate and cause disease in these mice. Notably, in the absence of PARP12, Mac1 mutant replication was increased in BMDMs and in mice. In addition, liver pathology was also increased in A59 infected mice. However, the PARP12 knockout did not restore Mac1 mutant virus replication to WT virus levels in all cell or tissue types and did not significantly increase the lethality of Mac1 mutant viruses. These results demonstrate that while PARP12 inhibits MHV Mac1 mutant virus infection, additional PARPs or innate immune factors must contribute to the extreme attenuation of this virus in mice. IMPORTANCEOver the last decade, the importance of ADP-ribosyltransferases (ARTs), also known as PARPs, in the antiviral response has gained increased significance as several were shown to either restrict virus replication or impact innate immune responses. However, there are few studies showing ART-mediated inhibition of virus replication or pathogenesis in animal models. We found that the CoV macrodomain (Mac1) was required to prevent ART-mediated inhibition of virus replication in cell culture. Here, using knockout mice, we found that PARP12, an interferon-stimulated ART, was required to repress the replication of a Mac1 mutant CoV both in cell culture and in mice, demonstrating that PARP12 represses coronavirus replication. However, the deletion of PARP12 did not fully rescue Mac1 mutant virus replication or pathogenesis, indicating that multiple PARPs function to counter coronavirus infection.

microbiology↗

bioRxiv2023

SARS-CoV-2 Mac1 is required for IFN antagonism and efficient virus replication in mice

Several coronavirus (CoV) encoded proteins are being evaluated as targets for antiviral therapies for COVID-19. Included in this set of proteins is the conserved macrodomain, or Mac1, an ADP-ribosylhydrolase and ADP-ribose binding protein. Utilizing point mutant recombinant viruses, Mac1 was shown to be critical for both murine hepatitis virus (MHV) and severe acute respiratory syndrome (SARS)-CoV virulence. However, as a potential drug target, it is imperative to understand how a complete Mac1 deletion impacts the replication and pathogenesis of different CoVs. To this end, we created recombinant bacterial artificial chromosomes (BACs) containing complete Mac1 deletions ({Delta}Mac1) in MHV, MERS-CoV, and SARS-CoV-2. While we were unable to recover infectious virus from MHV or MERS-CoV {Delta}Mac1 BACs, SARS-CoV-2 {Delta}Mac1 was readily recovered from BAC transfection, indicating a stark difference in the requirement for Mac1 between different CoVs. Furthermore, SARS-CoV-2 {Delta}Mac1 replicated at or near wild-type levels in multiple cell lines susceptible to infection. However, in a mouse model of severe infection, {Delta}Mac1 was quickly cleared causing minimal pathology without any morbidity. {Delta}Mac1 SARS-CoV-2 induced increased levels of interferon (IFN) and interferon-stimulated gene (ISG) expression in cell culture and mice, indicating that Mac1 blocks IFN responses which may contribute to its attenuation. {Delta}Mac1 infection also led to a stark reduction in inflammatory monocytes and neutrophils. These results demonstrate that Mac1 only minimally impacts SARS-CoV-2 replication, unlike MHV and MERS-CoV, but is required for SARS-CoV-2 pathogenesis and is a unique antiviral drug target. SIGNIFICANCEAll CoVs, including SARS-CoV-2, encode for a conserved macrodomain (Mac1) that counters host ADP-ribosylation. Prior studies with SARS-CoV-1 and MHV found that Mac1 blocks IFN production and promotes CoV pathogenesis, which has prompted the development of SARS-CoV-2 Mac1 inhibitors. However, development of these compounds into antivirals requires that we understand how SARS-CoV-2 lacking Mac1 replicates and causes disease in vitro and in vivo. Here we found that SARS-CoV-2 containing a complete Mac1 deletion replicates normally in cell culture but induces an elevated IFN response, has reduced viral loads in vivo, and does not cause significant disease in mice. These results will provide a roadmap for testing Mac1 inhibitors, help identify Mac1 functions, and open additional avenues for coronavirus therapies.

microbiology↗

bioRxiv2022

T-cell cellular stress and reticulocyte signatures, but not loss of naive T lymphocytes, characterize severe COVID-19 in older adults

In children and younger adults up to 39 years of age, SARS-CoV-2 usually elicits mild symptoms that resemble the common cold. Disease severity increases with age starting at 30 and reaches astounding mortality rates that are ~330 fold higher in persons above 85 years of age compared to those 18-39 years old. To understand age-specific immune pathobiology of COVID-19 we have analyzed soluble mediators, cellular phenotypes, and transcriptome from over 80 COVID-19 patients of varying ages and disease severity, carefully controlling for age as a variable. We found that reticulocyte numbers and peripheral blood transcriptional signatures robustly correlated with disease severity. By contrast, decreased numbers and proportion of naive T-cells, reported previously as a COVID-19 severity risk factor, were found to be general features of aging and not of COVID-19 severity, as they readily occurred in older participants experiencing only mild or no disease at all. Single-cell transcriptional signatures across age and severity groups showed that severe but not moderate/mild COVID-19 causes cell stress response in different T-cell populations, and some of that stress was unique to old severe participants, suggesting that in severe disease of older adults, these defenders of the organism may be disabled from performing immune protection. These findings shed new light on interactions between age and disease severity in COVID-19.

immunology↗

bioRxiv2022

Toroidal displacement of Klebsiella pneumoniae by Pseudomonas aeruginosa is a unique mechanism to avoid competition for iron

Competition for resources is one of the major drivers for evolution and retention of new traits in microbial communities. Quorum-dependent traits of opportunistic human pathogen Pseudomonas aeruginosa allow it to survive and thrive in nature. Here, we report a unique surfactant-driven pushing mechanism that P. aeruginosa employs specifically against Klebsiella pneumoniae. The pushing is accomplished in a manner that is dependent on nutrient limitation and quorum sensing. We find that P. aeruginosa employs neither proteases nor toxic secondary metabolites against K. pneumoniae. Rhamnolipid biosurfactant appears to be the only factor required to displace Klebsiella effectively. Both rhamnolipid production and the pushing ability of P. aeruginosa are suppressed by iron supplementation. We show that both these bacteria produce several siderophores in minimal medium and rapidly deplete iron. Under these conditions, P. aeruginosa pushes Klebsiella away from the substratum using rhamnolipid, reducing the competition for iron. Our study describes a unique quorum and iron-responsive mechanism in P. aeruginosa to support its own growth during resource competition.

microbiology↗

bioRxiv2022

Oxytocin receptor is not required for social attachment in prairie voles

Prairie voles are among a small group of mammals that display long-term social attachment between mating partners. Many pharmacological studies show that signaling via the oxytocin receptor (OxtR) is critical for the display of social monogamy in these animals. We used CRISPR-mutagenesis to independently generate three different OxtR null mutant prairie vole lines. OxtR mutants displayed social attachment such that males and females showed a behavioral preference for their mating partners over a stranger of the opposite sex when assayed using different paradigms. Mothers lacking OxtR delivered viable pups, and parents displayed care of their young and raised them to the weanling stage. Together, our studies unexpectedly reveal that OxtR-mediated signaling is genetically dispensable for social attachment, parturition, and parental behavior.

animal behavior and cognition↗

bioRxiv2022

Enriching Representation Learning Using 53 Million Patient Notes through Human Phenotype Ontology Embedding

The Human Phenotype Ontology (HPO) is a dictionary of more than 15,000 clinical phenotypic terms with defined semantic relationships, developed to standardize their representation for phenotypic analysis. Over the last decade, the HPO has been used to accelerate the implementation of precision medicine into clinical practice. In addition, recent research in representation learning, specifically in graph embedding, has led to notable progress in automated prediction via learned features. Here, we present a novel approach to phenotype representation by incorporating phenotypic frequencies based on 53 million full-text health care notes from more than 1.5 million individuals. We demonstrate the efficacy of our proposed phenotype embedding technique by comparing our work to existing phenotypic similarity-measuring methods. Using phenotype frequencies in our embedding technique, we are able to identify phenotypic similarities that surpass the current computational models. In addition, we show that our embedding technique aligns with domain experts judgment at a level that exceeds their agreement. We show that our proposed technique efficiently represents complex and multidimensional phenotypes in HPO format, which can then be used as input for various downstream tasks that require deep phenotyping, including patient similarity analyses and disease trajectory prediction.

bioinformatics↗

bioRxiv2022

The SCN1A Philadelphia variant - a gain-of-function mutation causing an early-onset epileptic encephalopathy

ObjectiveLoss-of-function variants in SCN1A cause Dravet Syndrome, the most common genetic developmental and epileptic encephalopathy (DEE). However, emerging evidence suggests separate entities of SCN1A-related disorders due to gain-of-function variants. Here, we aim to refine the clinical, genetic, and functional electrophysiological features of a recurrent p.R1636Q gain-of-function variant, identified in four individuals at a single center. MethodsIndividuals carrying the recurrent SCN1A p.R1636Q variant were identified through diagnostic testing. Whole-cell voltage-clamp electrophysiological recording in HEK-293T cells was performed to compare the properties of sodium channels containing wild-type Nav1.1 or Nav1.1-R1636Q along with both Nav{beta}1 and Nav{beta}2 subunits, including response to oxcarbazepine. To delineate differences to other SCN1A-related epilepsies, we analyzed electronic medical records. ResultsAll four individuals had an early-onset DEE characterized by focal tonic seizures and additional seizure types starting in the first few weeks of life. Electrophysiological analysis showed a mixed gain-of-function effect with normal current density, a leftward (hyperpolarized) shift of steady-state inactivation, and slower inactivation kinetics leading to a prominent late sodium current (INa). The observed functional changes closely paralleled effects of pathogenic variants in SCN3A and SCN8A at corresponding positions. Both wildtype and variant exhibited sensitivity to block by oxcarbazepine, partially correcting electrophysiological abnormalities of the SCN1A p.R1636Q variant. Clinically, a single individual responded to treatment with oxcarbazepine. Across 51 individuals with SCN1A-related epilepsies, those with the recurrent p.R1636Q variants had the earliest ages of onset. InterpretationThe recurrent SCN1A p.R1636Q variant causes a clinical entity with a wider clinical spectrum than previously reported, characterized by ultra early-onset epilepsy and absence of prominent movement disorder. Functional consequences of this variant lead to mixed loss- and gain-of-function that is partially corrected by oxcarbazepine. The recurrent p.R1636Q variant represents one of the most common causes of early-onset SCN1A-related epilepsies with separate treatment and prognosis implications. Key PointsO_LILoss-of-function variants in SCN1A cause Dravet syndrome, but gain-of-function variants have an emerging clinical spectrum. C_LIO_LIThe SCN1A p.R1636Q variant shows similar overall gain-of-function effects to identical missense variants in other voltage-gated sodium channels. C_LIO_LIFeatures of four unreported individuals with SCN1A p.R1636Q from a single center expand the SCN1A gain-of-function phenotype. C_LIO_LIIndividuals with this variant are recognizable by their ultra early-onset seizures in contrast to Dravet syndrome. C_LI

genetics↗

bioRxiv2022

A recurrent de novo splice site variant involving DNM1 alternative exon 10a causes developmental and epileptic encephalopathy through a dominant-negative mechanism

Heterozygous pathogenic variants in DNM1 cause developmental and epileptic encephalopathy (DEE) due to a dominant-negative mechanism impeding vesicular fission. Thus far, pathogenic variants in DNM1 have been studied using a canonical transcript that includes the alternatively spliced exon 10b. However, after performing RNA sequencing in thirty-nine pediatric brain samples, we find the primary transcript expressed in the brain includes the downstream exon 10a instead. Using this information, we evaluated genotype-phenotype correlations of variants affecting exon 10a and identified a cohort of eleven previously unreported individuals. Eight individuals harbor a recurrent de novo splice site variant, NG_029726.1(NM_001288739.1):c.1197-8G>A, which affects exon 10a and leads to DEE consistent with the classical DNM1 phenotype. We find this splice site variant leads to disease through an unexpected dominant-negative mechanism. Functional testing reveals an in-frame upstream splice acceptor causing insertion of two amino acids predicted to impair oligomerization-dependent activity. This is supported by neuropathological samples showing accumulation of synaptic vesicles adherent to the plasma membrane consistent with impaired vesicular fission. Two additional individuals with missense variants affecting exon 10a, p.(Arg399Trp) and p.(Gly401Asp), had a similar DEE phenotype. In contrast, a single individual with a missense variant affecting exon 10b, p.(Pro405Leu), which is less expressed in the brain, had a correspondingly less severe presentation. Thus, we implicate variants affecting exon 10a as causing the severe DEE typically associated with DNM1-related disorders. We highlight the importance of considering relevant isoforms for disease-causing variants, as well as the possibility of splice site variants acting through a dominant-negative mechanism.

genetics↗

bioRxiv2021

Predicting the functional effects of voltage-gated potassium channel missense variants with multi-task learning

PurposeVariants in genes encoding voltage-gated potassium channels are associated with a broad spectrum of neurological diseases including epilepsy, ataxia, and intellectual disability. Knowledge of the resulting functional changes, characterized as overall ion channel gain- or loss-of-function, is essential to guide clinical management including precision medicine therapies. However, for an increasing number of variants, little to no experimental data is available. New tools are needed to evaluate variant functional effects. MethodsWe catalogued a comprehensive dataset of 959 functional experiments across 19 voltage-gated potassium channels, leveraging data from 782 unique disease-associated and synthetic variants. We used these data to train a taxonomy-based multi-task learning support vector machine (MTL-SVM), and compared performance to a baseline of standard SVMs. ResultsMTL-SVM maintains channel family structure during model training, improving overall predictive performance (mean balanced accuracy 0.729 {+/-} 0.029, AU-ROC 0.757 {+/-} 0.039) over baseline (mean balanced accuracy 0.645 {+/-} 0.041, AU-ROC 0.710 {+/-} 0.074). We can obtain meaningful predictions even for channels with few known variants (KCNC1, KCNQ5). ConclusionOur model enables functional variant prediction for voltage-gated potassium channels. It may assist in tailoring current and future precision therapies for the increasing number of patients with ion channel disorders.

bioinformatics↗