Biology subjects

Marwaha, S.

Publications and source records attributed to Marwaha, S..

7 recordsLinked to original sources

bioRxiv2026

Building an Interoperable Rare Disease Multi-omic Resource: The GREGoR Data Model and Dataset

Rare disease research and diagnosis rely on the integration of genomic and phenotypic data generated across diverse clinical sites; however, the absence of widely adopted standards for representing genomic data and associated metadata has limited data interoperability, reuse, and cross-study analysis. The Genomics Research to Elucidate the Genetics of Rare Diseases (GREGoR) Consortium was established to investigate challenging rare disease cases and evaluate emerging multi-omic technologies for clinical translation. To support coordinated data integration across distributed research sites, we developed a common Consortium Data Model in partnership with domain experts to standardize the capture of participant-, family-, phenotype- and assay-level metadata, with a particular emphasis on using a modular architecture to support linking of multiple data versions from multiple omic technologies to a single individual and attribution of a genetic finding to the specific technology used for its initial discovery. Adoption of the GREGoR Data Model has enabled continued generation and public release of a harmonized, analysis-ready Consortium Dataset. The most recent release includes phenotypic, family and multi-omic data from 12,292 participants in 5,029 families. Other rare disease data sharing efforts are beginning to adopt this data model which will facilitate cross consortium analyses and empower rare disease research. This work demonstrates that a collaborative, flexible, and scalable data model can enable large-scale rare disease research, facilitate cross-center data harmonization, and enable data interoperability.

genomics↗

bioRxiv2025

Biomni: A General-Purpose Biomedical AI Agent

Biomedical research underpins progress in our understanding of human health and disease, drug discovery, and clinical care. However, with the growth of complex lab experiments, large datasets, many analytical tools, and expansive literature, biomedical research is increasingly constrained by repetitive and fragmented workflows that slow discovery and limit innovation, underscoring the need for a fundamentally new way to scale scientific expertise. Here, we introduce Biomni, a general-purpose biomedical AI agent designed to autonomously execute a wide spectrum of research tasks across diverse biomedical subfields. To systematically map the biomedical action space, Biomni first employs an action discovery agent to create the first unified agentic environment - mining essential tools, databases, and protocols from tens of thousands of publications across 25 biomedical domains. Built on this foundation, Biomni features a generalist agentic architecture that integrates large language model (LLM) reasoning with retrieval-augmented planning and code-based execution, enabling it to dynamically compose and carry out complex biomedical workflows - entirely without relying on predefined templates or rigid task flows. Systematic benchmarking demonstrates that Biomni achieves strong generalization across heterogeneous biomedical tasks - including causal gene prioritization, drug repurposing, rare disease diagnosis, micro-biome analysis, and molecular cloning - without any task-specific prompt tuning. Real-world case studies further showcase Biomnis ability to interpret complex, multi-modal biomedical datasets and autonomously generate experimentally testable protocols. Biomni envisions a future where virtual AI biologists operate alongside and augment human scientists to dramatically enhance research productivity, clinical insight, and healthcare. Biomni is ready to use at https://biomni.stanford.edu, and we invite scientists to explore its capabilities, stress-test its limits, and co-create the next era of biomedical discoveries.

bioinformatics↗

bioRxiv2023

RExPRT: a machine learning tool to predict pathogenicity of tandem repeat loci

Tandem repeats (TRs) are polymorphic sequences of DNA that are composed of repeating units of motifs, whose lengths can vary depending on the type of TR. Expansions of TRs are responsible for approximately 50 monogenic diseases, compared to over 4,300 disease causing genes disrupted by single nucleotide variants and small indels. It appears thus reasonable to expect the discovery of additional pathogenic repeat expansions, which has the potential of significantly narrowing the current diagnostic gap in many diseases. Recently, short and long-read whole genome sequencing with the use of advanced bioinformatics tools, have identified a growing number of TR expansions in the human population. The majority of these loci are expanded in <1% of genomes. Categorizing and prioritizing such TR loci is a growing challenge to human genomic studies. We present a first-in-class machine learning tool, RExPRT (Repeat EXpansion Pathogenicity pRediction Tool), which is designed to distinguish pathogenic from benign TR expansions. Leave-one-out cross validation results demonstrated that an ensemble approach comprised of support vector machines (SVM) and extreme gradient boosted decision tree (XGB) classify TRs with a precision of 92% and a recall of 90%. Further validation of RExPRT on unseen test data demonstrate a similar precision of 86%, and a recall of 60%. RExPRTs high precision in particular, will be of significant value to large-scale discovery studies, which require the prioritization of promising candidate loci for time-consuming and costly functional follow-up studies. Application of RExPRT to ~800,000 TRs in the reference genome identified ~30,000 TRs that would be likely pathogenic upon expansion. Thus, RExPRT establishes a foundation for the application of machine learning approaches to categorize the pathogenicity of tandem repeat loci.

bioinformatics↗

bioRxiv2023

The mitochondrial multi-omic response to exercise training across tissues

Mitochondria are adaptable organelles with diverse cellular functions critical to whole-body metabolic homeostasis. While chronic endurance exercise training is known to alter mitochondrial activity, these adaptations have not yet been systematically characterized. Here, the Molecular Transducers of Physical Activity Consortium (MoTrPAC) mapped the longitudinal, multi-omic changes in mitochondrial analytes across 19 tissues in male and female rats endurance trained for 1, 2, 4 or 8 weeks. Training elicited substantial changes in the adrenal gland, brown adipose, colon, heart and skeletal muscle, while we detected mild responses in the brain, lung, small intestine and testes. The colon response was characterized by non-linear dynamics that resulted in upregulation of mitochondrial function that was more prominent in females. Brown adipose and adrenal tissues were characterized by substantial downregulation of mitochondrial pathways. Training induced a previously unrecognized robust upregulation of mitochondrial protein abundance and acetylation in the liver, and a concomitant shift in lipid metabolism. The striated muscles demonstrated a highly coordinated response to increase oxidative capacity, with the majority of changes occurring in protein abundance and post-translational modifications. We identified exercise upregulated networks that are downregulated in human type 2 diabetes and liver cirrhosis. In both cases HSD17B10, a central dehydrogenase in multiple metabolic pathways and mitochondrial tRNA maturation, was the main hub. In summary, we provide a multi-omic, cross-tissue atlas of the mitochondrial response to training and identify candidates for prevention of disease-associated mitochondrial dysfunction.

molecular biology↗

bioRxiv2022

Temporal dynamics of the multi-omic response to endurance exercise training across tissues

Regular exercise promotes whole-body health and prevents disease, yet the underlying molecular mechanisms throughout a whole organism are incompletely understood. Here, the Molecular Transducers of Physical Activity Consortium (MoTrPAC) profiled the temporal transcriptome, proteome, metabolome, lipidome, phosphoproteome, acetylproteome, ubiquitylproteome, epigenome, and immunome in whole blood, plasma, and 18 solid tissues in Rattus norvegicus over 8 weeks of endurance exercise training. The resulting data compendium encompasses 9466 assays across 19 tissues, 25 molecular platforms, and 4 training time points in young adult male and female rats. We identified thousands of shared and tissue- and sex-specific molecular alterations. Temporal multi-omic and multi-tissue analyses demonstrated distinct patterns of tissue remodeling, with widespread regulation of immune, metabolism, heat shock stress response, and mitochondrial pathways. These patterns provide biological insights into the adaptive responses to endurance training over time. For example, exercise training induced heart remodeling via altered activity of the Mef2 family of transcription factors and tyrosine kinases. Translational analyses revealed changes that are consistent with human endurance training data and negatively correlated with disease, including increased phospholipids and decreased triacylglycerols in the liver. Sex differences in training adaptation were widespread, including those in the brain, adrenal gland, lung, and adipose tissue. Integrative analyses generated novel hypotheses of disease relevance, including candidate mechanisms that link training adaptation to non-alcoholic fatty liver disease, inflammatory bowel disease, cardiovascular health, and tissue injury and recovery. The data and analysis results presented in this study will serve as valuable resources for the broader community and are provided in an easily accessible public repository (https://motrpac-data.org/). HighlightsO_LIMulti-tissue resource identifies 35,439 analytes regulated by endurance exercise training at 5% FDR across 211 combinations of tissues and molecular platforms. C_LIO_LIInterpretation of systemic and tissue-specific molecular adaptations produced hypotheses to help describe the health benefits induced by exercise. C_LIO_LIRobust sex-specific responses to endurance exercise training are observed across multiple organs at the molecular level. C_LIO_LIDeep multi-omic profiling of six tissues defines regulatory signals for tissue adaptation to endurance exercise training. C_LIO_LIAll data are available in a public repository, and processed data, analysis results, and code to reproduce major analyses are additionally available in convenient R packages. C_LI

molecular biology↗

bioRxiv2019

Machine learning-based detection of insertions and deletions in the human genome

Insertions and deletions (indels) make a critical contribution to human genetic variation. While indel calling has improved significantly, it lags dramatically in performance relative to single-nucleotide variant calling, something of particular concern for clinical genomics where larger scale disruption of the open reading frame can commonly cause disease. Here, we present a machine learning-based approach to the detection of indel breakpoints called Scotch. This novel approach improves sensitivity to larger variants dramatically by leveraging sequencing metrics and signatures of poor read alignment. We also introduce a meta-analytic indel caller, called Metal, that performs a "smart intersection" of Scotch and currently available tools to be maximally sensitive to large variants. We use new benchmark datasets and Sanger sequencing to compare Scotch and Metal to current gold standard indel callers, achieving unprecedented levels of precision and recall. We demonstrate the impact of these improvements by applying this tool to a cohort of patients with undiagnosed disease, generating plausible novel candidates in 21 out of 26 undiagnosed cases. We highlight the diagnosis of one patient with a 498-bp deletion in HNRNPA1 missed by traditional indel-detection tools.

bioinformatics↗

bioRxiv2018

Identification of rare-disease genes in diverse undiagnosed cases using whole blood transcriptome sequencing and large control cohorts

RNA sequencing (RNA-seq) is a complementary approach for Mendelian disease diagnosis for patients in whom exome-sequencing is not informative. For both rare neuromuscular and mitochondrial disorders, its application has improved diagnostic rates. However, the generalizability of this approach to diverse Mendelian diseases has yet to be evaluated. We sequenced whole blood RNA from 56 cases with undiagnosed rare diseases spanning 11 diverse disease categories to evaluate the general application of RNA-seq to Mendelian disease diagnosis. We developed a robust approach to compare rare disease cases to existing large sets of RNA-seq controls (N=1,594 external and N=31 family-based controls) and demonstrated the substantial impacts of gene and variant filtering strategies on disease gene identification when combined with RNA-seq. Across our cohort, we observed that RNA-seq yields a 8.5% diagnostic rate. These diagnoses included diseases where blood would not intuitively reflect evidence of disease. We identified RARS2 as an under-expression outlier containing compound heterozygous pathogenic variants for an individual exhibiting profound global developmental delay, seizures, microcephaly, hypotonia, and progressive scoliosis. We also identified a new splicing junction in KCTD7 for an individual with global developmental delay, loss of milestones, tremors and seizures. Our study provides a broad evaluation of blood RNA-seq for the diagnosis of rare disease.

genomics↗