Biology subjects

Rubin, D.

Publications and source records attributed to Rubin, D..

7 recordsLinked to original sources

bioRxiv2026

A generalizable speech neuroprosthesis

Intracortical brain-computer interfaces (BCIs) can restore communication to people with vocal tract paralysis by decoding cortical activity during attempted speech into text. State-of-the-art systems pairing neural-to-phoneme decoders with phoneme-to-word language models have achieved word error rates (WERs) as low as 1%, but only after collecting thousands of sentences of training data. Shortening the data collection process would facilitate scaling this new technology by reducing the time from device implant to high-accuracy communication. Here we introduce a transformer-based decoder model trained jointly across six intracortical speech BCI participants. For every participant -- regardless of sex, disease etiology, or attempted speaking strategy -- a multi-user model decoded speech more accurately (over 50% lower relative WER on average) than models trained on individual users data. Notably, the multi-user model could be finetuned on fewer than 200 sentences from a held-out user to achieve a WER below 7%. These results reveal how to pool intracortical data across people to yield more accurate, generalizable, and rapidly-deployable decoding models.

bioengineering↗

bioRxiv2024

Diet outperforms microbial transplant to drive microbiome recovery post-antibiotics

High-fat, low-fiber Western-style diets (WD) induce microbiome dysbiosis characterized by reduced taxonomic diversity and metabolic breadth1,2, which in turn increases risk for a wide array of metabolic3-5, immune6 and systemic pathologies. Recent work has established that WD can impair microbiome resilience to acute perturbations like antibiotic treatment7,8, although we know little about the mechanism of impairment and the specific host consequences of prolonged post-antibiotic dysbiosis. Here, we characterize the trajectory by which the gut microbiome recovers its taxonomic and functional profile after antibiotic treatment in mice on regular chow (RC) and WD, and find that only mice on RC undergo a rapid successional process of recovery. Metabolic modeling indicates that RC diet promotes the development of syntrophic cross- feeding interactions, while on WD, a dominant taxon monopolizes readily available resources without releasing syntrophic byproducts. Intervention experiments reveal that an appropriate dietary resource environment is both necessary and sufficient for rapid and robust microbiome recovery, whereas microbial transplant is neither. Furthermore, prolonged post-antibiotic dysbiosis in mice on WD renders them susceptible to infection by the intestinal pathogen Salmonella enterica serovar Typhimurium. Our data challenge widespread enthusiasm for fecal microbiota transplant (FMT) as a strategy to address dysbiosis and demonstrate that specific dietary interventions are, at minimum, an essential prerequisite for effective FMT, and may afford a safer, more natural, and less invasive alternative to FMT.

microbiology↗

bioRxiv2023

Microbially-catalyzed conjugation of GABA and tyramine to bile acids

Bile acids (BAs) are cholesterol-derived molecules that aid in digestion and nutrient absorption, regulate host metabolic processes, and influence physiology of the gut microbiota. Both the host and its microbiome contribute to enzymatic modifications that shape the chemical diversity of BAs in the gut. Several bacterial species have been reported to conjugate standard amino acids to BAs, but it was not known if bacteria conjugate BAs to other amine classes. Here, we show that Bacteroides fragilis strain P207, isolated from a bacterial bloom in the J-pouch of a patient with ulcerative colitis (UC) pouchitis, conjugates standard amino acids and the neuroactive amines {gamma}-aminobutyric acid (GABA) and tyramine to deoxycholic acid. We extended this analysis to other human gut isolates and identified species that are competent to conjugate GABA and tyramine to primary and secondary BAs, and further identified diverse BA-GABA and BA-tyramine amides in human stool. A longitudinal metabolomic analysis of J-pouch contents of the patient from whom B. fragilis P207 was isolated revealed highly reduced levels of secondary bile acids and a shifting BA amide profile before, during, and after onset of pouchitis, including temporal changes in several BA-GABA amides. Treatment of pouchitis with ciprofloxacin was associated with a marked reduction of nearly all BA amides in the J-pouch. Our study expands the known repertoire of conjugated bile acids produced by bacteria to include BA conjugates to GABA and tyramine and demonstrates that these molecules are present in the human gut. ImportanceBile acids (BAs) are modified in multiple ways by host enzymes and the microbiota to produce a chemically diverse set of molecules that assist in the digestive process and impact many physiological functions. This study reports the discovery of bacteria isolated from the gut of human patients that conjugate the neuroactive amines, GABA and tyramine, to BAs and demonstrates that BA-GABA and BA-tyramine amides are present in the human gut. GABA and tyramine are common metabolic products of the gut microbiota and potent neuroactive molecules, and their conjugation to BAs may influence receptor-mediated regulatory mechanisms of humans and their gut microbes.

microbiology↗

bioRxiv2023

Multi-omics analysis of a Bacteroides fragilis isolate from an ulcerative colitis patient defines genetic determinants of fitness in bile

Bacteroides fragilis comprises 1-5% of the gut microbiota in healthy humans but can expand to >50% of the population in ulcerative colitis (UC) patients experiencing inflammation. The mechanisms underlying such microbial blooms are poorly understood, but the gut of UC patients has physicochemical features that differ from healthy patients and likely impact microbial physiology. For example, levels of the secondary bile acid deoxycholate (DC) are highly reduced in the ileoanal J-pouch of UC colectomy patients. We isolated a B. fragilis strain from a UC patient with pouch inflammation (i.e. pouchitis) and developed it as a genetic model system to identify genes and pathways that are regulated by DC and that impact B. fragilis fitness in DC and crude bile. Treatment of B. fragilis with a physiologically relevant concentration of DC reduced cell growth and remodeled transcription of one-quarter of the genome. DC strongly induced expression of chaperones and select transcriptional regulators and efflux systems and downregulated protein synthesis genes. Using a barcoded collection of {approx}50,000 unique insertional mutants, we further defined B. fragilis genes that contribute to fitness in media containing DC or crude bile. Genes impacting cell envelope functions including cardiolipin synthesis, cell surface glycosylation, and systems implicated in sodium-dependent bioenergetics were major bile acid fitness factors. As expected, there was limited overlap between transcriptionally regulated genes and genes that impacted fitness in bile when disrupted. Our study provides a genome-scale view of a B. fragilis bile response and genetic determinants of its fitness in DC and crude bile. ImportanceThe Gram-negative bacterium, Bacteroides fragilis, is a common member of the human gut microbiota that colonizes multiple host niches and can influence human physiology through a variety of mechanisms. Identification of genes that enable B. fragilis to grow across a range of host environments has been impeded in part by the relatively limited genetic tractability of this species. We have developed a high-throughput genetic resource for a B. fragilis strain isolated from a UC pouchitis patient. Bile acids limit microbial growth and are altered in abundance in UC pouches, where B. fragilis often blooms. Using this resource, we uncovered pathways and processes that impact B. fragilis fitness in bile and that may contribute to population expansions during bouts of gut inflammation.

microbiology↗

bioRxiv2021

Multiplexed Imaging Analysis of the Tumor-Immune Microenvironment Reveals Predictors of Outcome in Triple-Negative Breast Cancer

Triple-negative breast cancer, the poorest-prognosis breast cancer subtype, lacks clinically approved biomarkers for patient risk stratification and treatment management. Prior literature has shown that interrogation of the tumor-immune microenvironment may be a promising approach for the discovery of methods to fill these gaps. Recently developed high-dimensional tissue imaging technology, such as multiplexed ion beam imaging, provide spatial context to protein expression in the microenvironment, allowing in-depth characterization of cellular processes. We demonstrate that profiling the functional proteins involved in cell-to-cell interactions in the microenvironment can predict recurrence and overall survival. We highlight the immunological relevance of the immunoregulatory proteins PD-1, PD-L1, IDO, and Lag3 by tying interactions involving them to recurrence and survival. Multivariate analysis reveals that our methods provide additional prognostic information compared to clinical variables. In this work, we present a computational pipeline for the examination of the tumor-immune microenvironment using multiplexed ion-beam imaging that produces interpretable results, and is generalizable to other cancer types.

cancer biology↗

bioRxiv2019

Deep Active Lesion Segmentation

Lesion segmentation is an important problem in computer-assisted diagnosis that remains challenging due to the prevalence of low contrast, irregular boundaries that are unamenable to shape priors. We introduce Deep Active Lesion Segmentation (DALS), a fully automated segmentation framework that leverages the powerful nonlinear feature extraction abilities of fully Convolutional Neural Networks (CNNs) and the precise boundary delineation abilities of Active Contour Models (ACMs). Our DALS framework benefits from an improved level-set ACM formulation with a per-pixel-parameterized energy functional and a novel multiscale encoder-decoder CNN that learns an initialization probability map along with parameter maps for the ACM. We evaluate our lesion segmentation model on a new Multiorgan Lesion Segmentation (MLS) dataset that contains images of various organs, including brain, liver, and lung, across different imaging modalities--MR and CT. Our results demonstrate favorable performance compared to competing methods, especially for small training datasets.

bioengineering↗

bioRxiv2018

Locoregional Radiogenomic Models Capture Gene Expression Heterogeneity in Glioblastoma

Radiogenomics mapping noninvasively determines important relationships between the molecular genotype and imaging phenotype of various tumors, allowing advances in both clinical care and cancer research. While early work has shown its technical feasibility, here we extend radiogenomic mapping to a locoregional level that can account for the molecular heterogeneity of tumors. To achieve this, our data processing pipeline relies on three main steps: 1) the use of multi-omics data fusion to generate a set of 100 interpretable gene modules, 2) the use of patch-based image analysis (specifically of contrast-enhanced T1-weighted weighted MR images) combined with Generalized Linear Models (GLM) to establish potential links between module expressions and local MR signal, and 3) the use of expression heatmaps based on GLMs decision values to explore visualization of tumor molecular heterogeneity. The performance of the proposed approach was evaluated using a leave-one-patient-out crossvalidation method as well as a separate validation data set. The top performing models were based on a small set of 20 features and yielded Area Under the receiver operating characteristic Curve (AUC) above 0.65 on the validation cohort for eight modules. Next, we demonstrate the clinical and biological interpretation of four modules using molecular expression heatmaps superimposed on clinical radiographic images, showing the potential for assessing tumor molecular heterogeneity and the utility of this method for precision treatment in clinical decision making and imaging surveillance.

bioinformatics↗