Biology subjects

Sanders, J.

Publications and source records attributed to Sanders, J..

15 recordsLinked to original sources

bioRxiv2025

Puget predicts gene expression across cell types using sequence and 3D chromatin organization data

Gene expression is governed by both linear DNA sequence and three-dimensional (3D) chromatin architecture. Most gene expression prediction models rely on sequence alone, thereby failing to capture structural context and to generalize to unseen cell types. We present Puget, a deep learning model that predicts cell type-specific gene expression from sequence and Hi-C data, which captures 3D chromatin organization. Puget pairs pretrained sequence and Hi-C encoders with a lightweight transformer decoder. Using paired Hi-C/RNA-seq from 36 human and 4 mouse biosamples, we evaluate Pugets ability to generalize to held-out genes, held-out biosamples, and from human to mouse. Relative to a sequence-only baseline, Puget improves cross-biosample Pearson correlation by up to 25% on highly variable genes in training biosamples and, unlike the sequence-only model, generalizes to held-out biosamples and across species. In addition, in silico perturbation experiments show that Puget can prioritize experimentally validated enhancer-gene pairs. Together, these results highlight a generalizable approach for modeling gene expression from sequence and 3D chromatin organization.

bioinformatics↗

bioRxiv2025

A procedure for controlling the false discovery rate of de novopeptide sequencing

De novo sequencing is a powerful method for identifying peptides from mass spectrometry proteomics experiments without the use of a protein database. However, applications of de novo sequencing are currently severely limited by the lack of a reliable procedure for controlling the false discovery rate (FDR). Here, we introduce an FDR control procedure for the de novo setting which is at least as powerful as database search, give empirical evidence that it is statistically valid, and demonstrate its utility on a set of common de novo applications.

bioinformatics↗

bioRxiv2025

Improvements to Casanovo, a deep learning de novo peptidesequencer

Casanovo is a state-of-the-art deep learning model for de novo peptide sequencing from mass spectrometry proteomics data. Here we report on a series of enhancements to Casanovo, aimed at improving the interpretability of the scores assigned to predicted peptides, generalizing the software for use in database search, speeding up training and prediction runtimes, and providing workflows and visualization tools to facilitate adoption of Casanovo and interpretation of its results. Our goal is to make Casanovo accurate and easy to use for applications such as metaproteomics, antibody sequencing, immunopeptidomics, and discovery of novel peptide sequences in standard proteomics analyses. Casanovo is available as open source at https://github.com/Noble-Lab/casanovo.

molecular biology↗

bioRxiv2024

A transformer model for de novo sequencing of data-independent acquisition mass spectrometry data

A core computational challenge in the analysis of mass spectrometry data is the de novo sequencing problem, in which the generating amino acid sequence is inferred directly from an observed fragmentation spectrum without the use of a sequence database. Recently, deep learning models have made significant advances in de novo sequencing by learning from massive datasets of high-confidence labeled mass spectra. However, these methods are primarily designed for data-dependent acquisition (DDA) experiments. Over the past decade, the field of mass spectrometry has been moving toward using data-independent acquisition (DIA) protocols for the analysis of complex proteomic samples due to their superior specificity and reproducibility. Hence, we present a new de novo sequencing model called Cascadia, which uses a transformer architecture to handle the more complex data generated by DIA protocols. In comparisons with existing approaches for de novo sequencing of DIA data, Cascadia achieves state-of-the-art performance across a range of instruments and experimental protocols. Additionally, we demonstrate Cascadias ability to accurately discover de novo coding variants and peptides from the variable region of antibodies.

bioinformatics↗

bioRxiv2024

Accounting for digestion enzyme bias in Casanovo

A key parameter of any proteomics mass spectrometry experiment is the identity of the enzyme that is used to digest proteins in the sample into peptides. The Casanovo de novo sequencing model was trained using data that was generated with trypsin digestion; consequently, the model prefers to predict peptides that end with the amino acids "K" or "R." This bias is desirable when the Casanovo is used to analyze data that was also generated using trypsin but can be problematic if the data was generated using some other digestion enzyme. In this work, we modify Casanovo to take as input the identify of the digestion enzyme, alongside each observed spectrum. We then train Casanovo with data generated using several different restriction enzymes, and we demonstrate that the resulting model successfully learns to capture enzyme-specific behavior. However, we find, surprisingly, that this new model does not yield a significant improvement in sequencing accuracy relative to a model trained without the enzyme information but using the same training set. This observation may have important implications for future attempts to make use of experimental metadata in de novo sequencing models.

bioinformatics↗

bioRxiv2024

Tumor-expressed SPPL3 supports innate anti-tumor immune responses.

The development of an effective anti-tumor response relies on the synergistic actions of various immune cells that recognize tumor cells via distinct receptors. Tumors, however, often manipulate receptor-ligand interactions to evade recognition by the immune system. Recently, we highlighted the role of neolacto-series glycosphingolipids (nsGSLs), produced by the enzyme {beta}1,3-N-acetylglucosaminyltransferase 5 (B3GNT5), in tumor immune escape. We previously demonstrated that loss of signal peptidase like 3 (SPPL3), an inhibitor of B3GNT5, results in elevated levels of nsGSLs and impairs CD8 T cell activation. The impact of loss of SPPL3 and an elevated nsGSL profile in tumor cells on innate immune recognition remains to be elucidated. This study investigates the anti-tumor efficacy of neutrophils, NK cells, and {gamma}{delta} T cells on tumor cells lacking SPPL3. Our findings demonstrate that SPPL3-deficient target cells are less susceptible to trogocytosis by neutrophils and killing by NK cells and {gamma}{delta} T cells. Mechanistically, SPPL3 influences trogocytosis and {gamma}{delta} T cell instigated killing through modulation of nsGSL expression while SPPL3-mediated reduced killing by NK cells is nsGSL-independent. The nsGSL-dependent SPPL3 sensitivity depends on the proximity of surface receptor domains to the cell membrane and the affinity of receptor-ligand interactions as shown with various sets of defined antibodies. Thus, SPPL3 expression by tumor cells alters crosstalk with immune cells through the receptor-ligand interactome thereby driving escape not only from adaptive but also from innate immunity. These data underline the importance of investigating a potential synergism of GSL synthesis inhibitors with current immune cell activating immunotherapies.

immunology↗

bioRxiv2024

A learned score function improves the power of mass spectrometry database search

One of the core problems in the analysis of protein tandem mass spectrometry data is the peptide assignment problem: determining, for each observed spectrum, the peptide sequence that was responsible for generating the spectrum. Two primary classes of methods are used to solve this problem: database search and de novo peptide sequencing. State-of-the-art methods for de novo sequencing employ machine learning methods, whereas most database search engines use hand-designed score functions to evaluate the quality of a match between an observed spectrum and a candidate peptide from the database. We hypothesize that machine learning models for de novo sequencing implicitly learn a score function that captures the relationship between peptides and spectra, and thus may be re-purposed as a score function for database search. Because this score function is trained from massive amounts of mass spectrometry data, it could potentially outperform existing, hand-designed database search tools. To test this hypothesis, we re-engineered Casanovo, which has been shown to provide state-of-the-art de novo sequencing capabilities, to assign scores to given peptide-spectrum pairs. We then evaluated the statistical power of this Casanovo score function, Casanovo-DB, to detect peptides on a benchmark of three mass spectrometry runs from three different species. Our results show that, at a 1% peptide-level false discovery rate threshold, Casanovo-DB outperforms existing hand-designed score functions by 35% to 88%. In addition, we show that re-scoring with the Percolator post-processor benefits Casanovo-DB more than other score functions, further increasing the number of detected peptides.

bioinformatics↗

bioRxiv2022

Multi-state occupancy model estimates probability of detection of an aquatic parasite using environmental DNA: Pseudoloma neurophilia in zebrafish aquaria

Detecting the presence of important parasites within a host and its environment is critical to understanding the dynamics that influence a pathogens ability to persist, while accurate detection is also essential for implementation of effective control strategies. Pseudoloma neurophilia is the most common pathogen reported in zebrafish (Danio rerio) research facilities. The only assays currently available for P. neurophilia, are through lethal sampling, often requiring euthanasia of the entire population for accurate estimates of prevalence in small populations. We present a non-lethal screening method to detect Pseudoloma neurophilia in tank water based on detection of environmental DNA (eDNA) from this microsporidum, using a previously developed qPCR assay that was adapted to the digital PCR (dPCR) platform. Using the generated dPCR data, a multi-state occupancy model was also implemented to predict the probability of detection in tank water under different flow regimes and pathogen prevalence. The occupancy model revealed that samples collected in static conditions were more informative than samples collected from flow-through conditions, with a probability of detection at 80% and 47%, respectively. There was also a positive correlation with the prevalence of infection in water and prevalence in fish based on qPCR.

microbiology↗

bioRxiv2021

UNIGEMS: plasmids and parts to facilitate teaching on assembly, gene expression control and logic in E. coli

Synthetic biology is as an excellent vehicle for education, as it enables creative combination of engineering and molecular biology approaches for quantitative characterisations of the assembled constructs. However, there is a limited number of resources available for such applications in the educational context, where straightforward setup, easily measurable phenotypes and extensibility are of particular importance. To expand the availability of education-friendly resources to teach synthetic biology and genetic engineering, we developed Unigems, a set of 10 plasmids that enable out-of-the-box investigations of principles of gene expression control, as well as more complex designs a biological logic gate. The system uses a common high-copy plasmid backbone and a common set of primers to enable Gibson-assembly of PCR-generated or synthesised parts into a target vector. It currently has two reporter genes with either two constitutive (high- or low-level) or two inducible (lactose- or arabinose-) promoters, as well as a single-plasmid implementation of an AND logic gate. The Unigems system has already been employed in undergraduate teaching settings, during outreach events and for training of iGEM teams. All plasmids have been deposited in Addgene.

synthetic biology↗

bioRxiv2021

A subset of CB002 xanthine analogues bypass p53-signaling to restore a p53 transcriptome and target an S-phase cell cycle checkpoint in tumors with mutated-p53

Mutations in TP53 occur commonly in the majority of human tumors and confer aggressive tumor phenotypes including metastasis and therapy resistance. CB002 and structural-analogues restore p53 signaling in tumors with mutant-p53 but we find that unlike other xanthines such as caffeine, pentoxifylline, and theophylline, they do not deregulate the G2-checkpoint. Novel CB002-analogues induce pro-apoptotic Noxa protein in an ATF3/4-dependent manner, whereas caffeine, pentoxifylline, and theophylline do not. By contrast to caffeine, CB002-analogues target an S-phase checkpoint associated with increased p-RPA/RPA2, p-ATR, decreased Cyclin A, p-histone H3 expression and downregulation of essential proteins in DNA-synthesis and -repair. CB002-analogue #4 enhances cell death, and decreases Ki-67 in patient-derived tumor-organoids without toxicity to normal human cells. Preliminary in vivo studies demonstrate anti-tumor efficacy in mice. Thus, a novel class of anti-cancer drugs show activation of p53 pathway signaling in tumors with mutated p53, and target an S-phase checkpoint.

cancer biology↗

bioRxiv2020

SUN-MKL1 crosstalk regulates nuclear deformation and fast motility of breast carcinoma cells in fibrillar ECM micro-environment

Aligned collagen fibers provide topography for the rapid migration of single tumor cells (streaming migration) to invade the surrounding stroma, move within tumor nests towards blood vessels to intravasate and form distant metastases. Mechanisms of tumor cell motility have been studied extensively in the 2D context, but the mechanistic understanding of rapid single tumor cell motility in the in vivo context is still lacking. Here, we show that streaming tumor cells in vivo use collagen fibers with diameters below 3 m. Employing 1D migration assays with matching in vivo fiber dimensions, we found a dependence of tumor cell motility on 1D substrate width, with cells moving the fastest and the most persistently on the narrowest 1D fibers (700 nm - 2.5 m). Interestingly, we also observed nuclear deformation in the absence of restricting extracellular matrix pores during high speed carcinoma cell migration in 1D, similar to the nuclear deformation observed in tumor cells in vivo. Further, we found that actomyosin machinery is aligned along the 1D axis and actomyosin contractility synchronously regulates cell motility and nuclear deformation. To further investigate the link between cell speed and nuclear deformation, we focused on the Linker of Nucleoskeleton and Cytoskeleton (LINC) complex proteins and SRF-MKL1 signaling, key regulators of mechanotransduction, actomyosin contractility and actin-based cell motility. Analysis of The Cancer Genome Atlas dataset showed a dramatic decrease in the LINC complex proteins SUN1 and SUN2 in primary tumor compared to the normal tissue. Disruption of LINC complex by SUN1+2 KD led to multi-lobular elongated nuclei, increased tumor cell motility and concomitant increase in F-actin, without affecting Lamin proteins. Mechanistically, we found that MKL1, an effector of changes in cellular G-actin to F-actin ratio, is required for increased 1D motility seen in SUN1+2 KD cells. Thus, we demonstrate a previously unrecognized crosstalk between SUN proteins and MKL1 transcription factor in modulating nuclear shape and carcinoma cell motility in an in vivo relevant 1D microenvironment.

cancer biology↗

bioRxiv2020

HiCRep.py: Fast comparison of Hi-C contact matrices in Python

Hi-C is the most widely used assay for investigating genome-wide 3D organization of chromatin. When working with Hi-C data, it is often useful to calculate the similarity between contact matrices in order to asses experimental reproducibility or to quantify relationships among Hi-C data from related samples. The HiCRep algorithm has been widely adopted for this task, but the existing R implementation suffers from run time limitations on high resolution Hi-C data or on large single-cell Hi-C datasets. We introduce a Python implementation of HiCRep and demonstrate that it is much faster than the existing R implementation. Furthermore, we give examples of HiCReps ability to accurately distinguish replicates from non-replicates and to reveal cell type structure among collections of Hi-C data. HiCRep.py and its documentation are available with a GPL license at https://github.com/Noble-Lab/hicrep. The software may be installed automatically using the pip package installer.

bioinformatics↗

bioRxiv2020

A standardized and reproducible method to measure decision-making in mice

Progress in science requires standardized assays whose results can be readily shared, compared, and reproduced across laboratories. Reproducibility, however, has been a concern in neuroscience, particularly for measurements of mouse behavior. Here we show that a standardized task to probe decision-making in mice produces reproducible results across multiple laboratories. We designed a task for head-fixed mice that combines established assays of perceptual and value-based decision making, and we standardized training protocol and experimental hardware, software, and procedures. We trained 140 mice across seven laboratories in three countries, and we collected 5 million mouse choices into a publicly available database. Learning speed was variable across mice and laboratories, but once training was complete there were no significant differences in behavior across laboratories. Mice in different laboratories adopted similar reliance on visual stimuli, on past successes and failures, and on estimates of stimulus prior probability to guide their choices. These results reveal that a complex mouse behavior can be successfully reproduced across multiple laboratories. They establish a standard for reproducible rodent behavior, and provide an unprecedented dataset and open-access tools to study decision-making in mice. More generally, they indicate a path towards achieving reproducibility in neuroscience through collaborative open-science approaches.

neuroscience↗

bioRxiv2019

Quantifying and understanding well-to-well contamination in microbiome research

Microbial sequences inferred as belonging to one sample may not have originated from that sample. Such contamination may arise from laboratory or reagent sources or from physical exchange between samples. This study seeks to rigorously assess the behavior of this often-neglected between-sample contamination. Using unique bacteria each assigned a particular well in a plate, we assess the frequency at which sequences from each source appears in other wells. We evaluate the effects of different DNA extraction methods performed in two labs using a consistent plate layout including blanks, low biomass, and high biomass samples. Well-to-well contamination occurred primarily during DNA extraction, and to a lesser extent in library preparation, while barcode leakage was negligible. Labs differed in the levels of contamination. DNA extraction methods differed in their occurrences and levels of well-to-well contamination, with robotic methods having more well-to-well contamination while manual methods having higher background contaminants. Well-to-well contamination was observed to occur primarily in neighboring samples, with rare events up to 10 wells apart. The effect of well-to-well was greatest in samples with lower biomass, and negatively impacted metrics of alpha and beta diversity. Our work emphasizes that sample contamination is a combination of crosstalk from nearby wells and background contaminants. To reduce well-to-well effects, samples should be randomized across plates, and samples of similar biomass processed together. Researchers should evaluate well-to-well contamination in study design and avoid removal of taxa or OTUs appearing in negative controls, as many will be microbes from other samples rather than reagent contaminants.\n\nImportanceMicrobiome research has uncovered magnificent biological and chemical stories across nearly all areas of life science, at times creating controversy when findings reveal fantastic descriptions of microbes living and even thriving in once thought to be sterile environments. Scientists have refuted many of these claims because of contamination, which has led to robust requirements including use of controls for validating accurate portrayals of microbial communities. In this study, we describe a previously undocumented form of contamination, well-to-well contamination and show that contamination primarily occurs during DNA extraction rather than PCR, is highest in plate-based methods as compared to single tube extraction, and occurs in higher frequency in low biomass samples. This finding has profound importance on the field as many current techniques to decontaminate a dataset simply relies on an assumption that microbial reads found in blanks are contaminants from outside namely the reagents or consumables.

genomics↗

bioRxiv2017

Multiethnic Meta-analysis Identifies New Loci for Pulmonary Function

Nearly 100 loci have been identified for pulmonary function, almost exclusively in studies of European ancestry populations. We extend previous research by meta-analyzing genome-wide association studies of 1000 Genomes imputed variants in relation to pulmonary function in a multiethnic population of 90,715 individuals of European (N=60,552), African (N=8,429), Asian (N=9,959), and Hispanic/Latino (N=11,775) ethnicities. We identified over 50 novel loci at genome-wide significance in ancestry-specific and/or multiethnic meta-analyses. Recent fine mapping methods incorporating functional annotation, gene expression, and/or differences in linkage disequilibrium between ethnicities identified potential causal variants and genes at known and newly identified loci. Sixteen of the novel genes encode proteins with predicted or established drug targets, including KCNK2 and CDK12.

genetics↗