Biology subjects

Gunawan, R.

Publications and source records attributed to Gunawan, R..

At least 19 recordsLinked to original sources

bioRxiv2025

SCREAM: Single-cell Clustering using Representation Autoencoder of Multiomics

MotivationSingle-cell multiomics technologies offer unprecedented opportunities to study cellular heterogeneity. But, integrating information across different omics modalities remains a major challenge due to high dimensionality, sparsity, and modality-specific noise characteristics. To address this, we develop SCREAM (Single-cell Clustering using Representation Autoencoder of Multiomics), a novel deep learning framework for the robust integration and clustering of multi-modal single-cell data. SCREAM leverages Stacked Autoencoders (SAEs) to generate robust latent representations for each omics modality as well as for their fusion. Subsequently, borrowing Deep Embedding Clustering (DEC), SCREAM iteratively fine tunes the integrated mulitomics latent space and single-cell cluster assignments. ResultsWe evaluated SCREAM against eleven state-of-the-art methods using SNARE-seq and CITE-seq datasets. In this benchmarking, SCREAM consistently demonstrated superior performance, yielding the highest or near-highest Adjusted Rand Index (ARI) and Normalized Mutual Information (NMI) scores on both datasets. These findings validate SCREAM as a highly accurate and robust approach for identifying cell types from multiomics data. Furthermore, its multiomics embeddings provides biologically meaningful latent representations for diverse downstream analyses. AvailabilitySCREAM is available at http://www.github.com/cabsel/scream. Contactrgunawan@buffalo.edu

bioinformatics↗

bioRxiv2025

GlycoEnzDB: A database of enzymes involved in human glycosylation

The glycan distribution on cells is governed by the stochastic activity of different families of enzymes that are together called glycoEnzymes. These include ~400 gene products or 2% of the proteome, that have recently been curated in an ontology called GlycoEnzOnto. With the goal of making this ontology more accessible to the larger biomedical and biotechnology community, we organized a web resource called GlycoEnzDB, presenting this enzyme classification both in terms of enzyme function and the pathways that they participate in. This information is linked to i) Figures from the "Essentials of Glycobiology" textbook, ii) General gene, enzyme and pathway data appearing in external databases, iii) Manual and generative-artificial intelligence (AI) based text describing the function and pathways regulated by these entities, iv) Single-cell expression data across cell lines, normal human cell-types and tissue, and v) CRISPR-knockout/activation/inactivation and Transcription factor activity predictions. Whereas these data are curated for human glycoEnzymes, the knowledge framework may be extended to other species also. The user-friendly web interface is accessible at www.virtualglycome.org/glycoenzdb.

biochemistry↗

bioRxiv2025

Transformer-based Deep Learning for Glycan Structure Inference from Tandem Mass Spectrometry

Glycans play critical roles in diverse biological processes, but their structural analysis by tandem mass spectrometry (MS/MS) remains a major challenge due to their branched structure and stereochemistry. Traditional computational methods, such as database searching, are constrained by the scope of existing libraries and can be computationally intensive. While recent deep learning models have advanced the field, they often struggle to capture the complex, long-range dependencies within MS/MS spectra required for accurate inference. To address these challenges, we present GlycoBERT and GlycoBART, novel transformer-based models for glycan structure prediction from MS/MS data. GlycoBERT, a sequence classifier, achieves 95.1% structural accuracy on test data, surpassing the current state-of-the-art deep learning model, CandyCrunch. However, classification-based methods are inherently limited to predicting structures present in the training data. To overcome this, we developed GlycoBART, a generative sequence-to-sequence model capable of de novo glycan inference. On independent validation datasets, both models demonstrate robust performance. Critically, when applied to an MS/MS dataset from human embryonic kidney cells, GlycoBART generated two de novo glycan structures absent from the training set, one of which is a novel structure not catalogued in major glycan databases. Together, GlycoBERT and GlycoBART establish a new benchmark for glycan analysis, offering a powerful framework that enables more accurate and comprehensive exploration of glycan diversity and discovery.

bioinformatics↗

bioRxiv2023

Cell and tissue-specific glycosylation pathways and transcriptional regulation informed by single-cell transcriptomics

While single cell studies have made significant impacts in various subfields of biology, they lag in the Glycosciences. To address this gap, we analyzed single-cell glycogene expressions in the Tabula Sapiens dataset of human tissues and cell types using a recent glycosylation-specific gene ontology (GlycoEnzOnto). At the median sequencing (count) depth, [~]40-50 out of 400 glycogenes were detected in individual cells. Upon increasing the sequencing depth, the number of detectable glycogenes saturates at [~]200 glycogenes, suggesting that the average human cell expresses about half of the glycogene repertoire. Hierarchies in glycogene and glycopathway expressions emerged from our analysis: nucleotide-sugar synthesis and transport exhibited the highest gene expressions, followed by genes for core enzymes, glycan modification and extensions, and finally terminal modifications. Interestingly, the same cell types showed variable glycopathway expressions based on their organ or tissue origin, suggesting nuanced cell- and tissue-specific glycosylation patterns. Probing deeper into the transcription factors (TFs) of glycogenes, we identified distinct groupings of TFs controlling different aspects of glycosylation: core biosynthesis, terminal modifications, etc. We present webtools to explore the interconnections across glycogenes, glycopathways, and TFs regulating glycosylation in human cell/tissue types. Overall, the study presents an overview of glycosylation across multiple human organ systems.

bioinformatics↗

bioRxiv2023

CrossTx: Cross-cell line Transcriptomic Signature Predictions

MotivationPredicting the cell response to chemical compounds is central to drug discovery, drug repurposing, and personalized medicine. To this end, large datasets of drug response signatures have been curated, most notably the Connectivity Map (CMap) from the Library of Integrated Network-based Cellular Signatures (LINCS) project. A multitude of in silico approaches have also been formulated to leverage drug signature data for accelerating novel therapeutics. However, the majority of the available data are from immortalized cancer cell lines. Cancer cells display markedly different responses to compounds, not only when compared to normal cells, but also among cancer types. Strategies for predicting drug signatures in unseen cells--cell lines not in the reference datasets--are still lacking. ResultsIn this work we developed a computational strategy, called CrossTx, for predicting drug transcriptomic signatures of an unseen target cell line using drug transcriptome data of reference cell lines and background transcriptome data of the target cells. Our strategy involves the combination of predictor and corrector steps. Briefly, the Predictor applies averaging (mean) or linear regression model to the reference dataset to generate cell line-agnostic drug signatures. The Corrector generates target-specific drug signatures by projecting cell line-agnostic signatures from the Predictor onto the transcriptomic latent space of the target cell line using Principal Component Analysis (PCA) and/or an Autoencoder (AE). We tested different combinations of Predictor-Corrector algorithms in an application to the CMap dataset to demonstrate the performance of our approach. ConclusionCrossTx is an efficacious and generalizable method for predicting drug signatures in an unseen target cell line. Among the combinations tested, we found that the best strategy is to employ Mean as the Predictor and PCA followed by AE (PCA+AE) as the Corrector. Still, the combination of Mean and PCA (without AE) is an attractive strategy because of its computationally efficiency and simplicity, while offering only slightly less accurate drug signature predictions than the best performing combination. Availability and implementationhttp://www.github.com/cabsel/crosstx Contactrgunawan@buffalo.edu

bioinformatics↗

bioRxiv2022

Dynamic changes in the niche and transcription trigger early murine and human pluripotent stem cell-derived liver organogenesis

The shift from collective migration to differentiation is a crucial process in epithelial biology but recreating this intricate transition has thus far proved elusive. We provide experimental, mechanistic, in vivo, and bioinformatic data supporting an undoubtable link between human pluripotent stem cell (hPSC)- derived collectively migrating hepatoblasts (MHB), and transcriptionally mature, functional hPSC- hepatocytes (HEPs), which incorporates two unrecognized steps. The protocol induces FOXA-dependent induction of HBs, leading to TBX3-positive, YAP-TEAD active MHBs which provide a transcriptional match with murine liver E9.5 MHBs. Simple cultivation changes trigger MHBs to rapidly form functional day 18 HEPs, predicted by a deep-learning designed gene circuit, resulting in a [~]236% fold- increase in maturation (PACNet), on par with the highest score, but with enhanced global transcriptional shaping. Overall, incorporating the MHB to HEP transition establishes a new, unrecognized, and highly efficient mechanism for differentiation that can be cumulatively integrated with existing methods to overcome barriers to maturation.

bioengineering↗

bioRxiv2022

Rewiring Dynamics of Functional Connectome in Motor Cortex during Motor Skill Learning

The brains connectome continually rewires throughout the life of an organism. In this study, we sought to elucidate the operational principles of such rewiring by analyzing the functional connectomes in mouse primary motor cortex (M1) during a 14-session (day) lever-press task learning in response to an auditory cue. Specifically, we employed Calcium imaging recordings of L2/3 and L5 of M1 in awake mice to reconstruct and analyze functional connectomes across learning sessions. Our results show that functional connectomes in L2/3 and L5 follow a similar learning-induced rewiring trajectory. More specifically, the connectomes rewire in a biphasic manner, where functional connectivity increases over the first few learning sessions, and then, it is gradually pruned to return to a homeostatic level of network density. We demonstrated that the increase of network connectivity in L2/3 connectomes, but not in L5, generates neuronal co-firing activity that correlates with higher motor performance (shorter cue-to-reward time), while motor performance remains relatively stable throughout the pruning phase. The results show a biphasic rewiring principle that involves the maximization of reward / performance and maintenance of network density. Finally, we demonstrated that the connectome rewiring in L2/3 is clustered around a core set of movement-associated neurons that form a highly interconnected hub in the connectomes, and that the activity of these core neurons stably encodes movement throughout learning. Significance StatementConnectomes in the motor cortex rewire during motor skill learning, but the operational principle behind this rewiring is yet to be determined. Here, we characterized the rewiring dynamics of functional connectomes in L2/3 and L5 of M1 in mice engaging in a lever-press learning, using two-photon fluorescence microscopy data. We identified a universal biphasic rewiring trajectory across animals and layers in the motor cortex that reflects two objectives: an exploratory phase that increases functional connectivity and optimizes motor performance, and a pruning phase that brings connectivity back to a homeostatic level while maintaining motor performance. We found further that connectome rewiring during motor skill learning concentrates around a core set of highly interconnected neurons in L2/3 that reliably encode movements.

neuroscience↗

bioRxiv2022

GlycoEnzOnto: A GlycoEnzyme Pathway and Molecular Function Ontology

The glycoEnzymes include a set of proteins having related enzymatic, metabolic, transport, structural and cofactor functions. Current there is no established ontology to describe glycoEnzyme properties and to relate them to glycan biosynthesis pathways. We present GlycoEnzOnto, an ontology describing 386 human glycoEnzymes curated along 135 glycosylation pathways, 134 molecular functions and 22 cellular compartments. The pathways described regulate nucleotide-sugar metabolism, glycosyl-substrate/donor transport, glycan biosynthesis, and degradation. The role of each enzyme in the glycosylation initiation, elongation/branching, and capping/termination phases is described. IUPAC linear strings present systematic human/machine readable descriptions of individual reaction steps and enable automated knowledge-based curation of biochemical networks. All GlycoEnzOnto knowledge is integrated with the Gene Ontology (GO) biological processes. GlycoEnzOnto enables improved transcript overrepresentation analyses and glycosylation pathway identification compared to other available schema, e.g. KEGG and Reactome. Overall, GlycoEnzOnto represents a holistic glycoinformatics resource for systems-level analyses. Availabilityhttps://github.com/neel-lab/GlycoEnzOnto

bioinformatics↗

bioRxiv2021

ΔFBA: Predicting metabolic flux alterations using genome-scale metabolic models and differential transcriptomic data

Genome-scale metabolic models (GEMs) provide a powerful framework for simulating the entire set of biochemical reactions occurring in a cell. Constraint-based modeling tools like flux balance analysis (FBA) developed for the purposes of predicting metabolic flux distribution using GEMs face considerable difficulties in estimating metabolic flux alterations between experimental conditions. Particularly, the most appropriate metabolic objective for FBA is not always obvious, likely context-specific, and not necessarily the same between conditions. Here, we propose a new method, called {Delta}FBA (deltaFBA), that employs constraint-based modeling, in combination with differential gene expression data, to evaluate changes in the intracellular flux distribution between two conditions. Notably, {Delta}FBA does not require specifying the cellular objective to produce the flux change predictions. We showcased the performance of {Delta}FBA through several case studies involving the prediction of metabolic alterations caused by genetic and environmental perturbations in Escherichia coli and caused by Type-2 diabetes in human muscle.

systems biology↗

bioRxiv2020

Nutrient signaling pathways regulate amyloid clearance and synaptic loss in Alzheimer's disease

Extra-cellular accumulation of Amyloid-{beta} (A{beta}) plaques is causatively associated with Alzheimers disease (AD). However, mechanisms that mediate the pre-pathological state of amyloid plaque formation remain elusive. Here, using paired RNAi and kinase inhibitor screens, we discovered that AKT-mediated insulin/nutrient signaling suppresses lysosomal clearance of A{beta} and promotes amyloid formation. This mechanism is cell-autonomous and functions in multiple systems, including iPSC-derived human neurons and in vivo. Nutrient signaling regulates amyloid formation via distinct lysosomal functional mechanisms, while enhanced amino acid signaling promotes amyloid formation by transcriptionally suppressing lysosome biogenesis, and high intracellular cholesterol levels suppress lysosomal clearance of amyloid by increasing the number of non-functional lysosomes. The nutrient signaling pathway, present in both neurons and microglia, regulates lysosomal clearance of amyloid and microglia mediated synapse loss, both in vitro and in vivo. Clinically, older hyperlipidemic patients showed less synapse loss through microglia and performed better in cognitive tests. Thus, our results reveal a bi-partite cellular quality control system regulated by the insulinnutrient signaling that in neurons regulates A{beta} peptide clearance and in microglia regulates synaptic loss, both processes causally associated with AD. Our results also caution against reducing amyloid through such processes as this might also result in synapse loss.

neuroscience↗

bioRxiv2020

FARCI: Fast and Robust Connectome Inference

The inference of neuronal connectome from large-scale neuronal activity recordings, such as two-photon Calcium imaging, represents an active area of research in computational neuroscience. In this work, we developed FARCI (Fast and Robust Connectome Inference), a MATLAB package for neuronal connectome inference from high-dimensional two-photon Calcium fluorescence data. We employed partial correlations as a measure of the functional association strength between pairs of neurons to reconstruct a neuronal connectome. We demonstrated using gold standard datasets from the Neural Connectomics Challenge (NCC) that FARCI provides an accurate connectome and its performance is robust to network sizes, missing neurons, and noise levels. Moreover, FARCI is computationally efficient and highly scalable to large networks. In comparison to the best performing algorithm in the NCC, FARCI produces more accurate networks over different network sizes and subsampling, while providing over two orders of magnitude faster computational speed.

neuroscience↗

bioRxiv2020

Selective silencing rather than targeted activation of gene expression underlies fate choice in human hematopoietic stem cells.

When human cord blood derived CD34+ cells are induced to differentiate in vitro, they undergo rapid and dynamic morphological and molecular transformations that are critical for fate commitment. Using ATAC-seq and single-cell RNA sequencing, we detected two phases in this process. In the first phase, we observed a rapid and global chromatin opening that makes most of the gene promoters in the genome accessible, followed by widespread upregulation of gene transcription and a concomitant increase in the cell-to-cell variability of gene expression. The second phase is marked by a slow chromatin closure and a subsequent overall downregulation of gene transcription and emergence of coherent expression profiles corresponding to distinct cell subpopulations. These observations are consistent with a model based on the spontaneous probabilistic organization of the cellular process of fate commitment.

cell biology↗

bioRxiv2020

Universality of cell differentiation trajectories revealed by a reconstruction of transcriptional uncertainty landscapes from single-cell transcriptomic data

We employed our previously-described single-cell gene expression analysis CALISTA (Clustering And Lineage Inference in Single-Cell Transcriptional Analysis) to evaluate transcriptional uncertainty at the single-cell level using a stochastic mechanistic model of gene expression. We reconstructed a transcriptional uncertainty landscape during cell differentiation by visualizing single-cell transcriptional uncertainty surface over a two dimensional representation of the single-cell gene expression data. The reconstruction of transcriptional uncertainty landscapes for ten publicly available single-cell gene expression datasets from cell differentiation processes with linear, single or multi-branching cell lineage, reveals universal features in the cell differentiation trajectory that include: (i) a peak in single-cell uncertainty during transition states, and in systems with bifurcating differentiation trajectories, each branching point represents a state of high transcriptional uncertainty; (ii) a positive correlation of transcriptional uncertainty with transcriptional burst size and frequency; (iii) an increase in RNA velocity preceeding the increase in the cell transcriptional uncertainty. Finally, we provided biological interpretations of the universal rise-then-fall profile of the transcriptional uncertainty landscape, including a link with the Waddingtons epigenetic landscape, that is generalizable to every cell differentiation system.

bioinformatics↗

bioRxiv2019

Heterogeneity in heat shock response dynamics caused by translation fidelity decline and proteostasis collapse

Genetics, environment, and stochasticity influence the rate of ageing in living organisms. Individual Caenorhabditis elegans that are genetically identical and cultured in the same environment have different lifespans, suggesting a significant role of stochasticity in ageing. We have developed a novel microfluidic methodology to measure heat-shock response as a surrogate marker for heterogeneity associated with lifespan and have quantified the heat-shock response of C. elegans at the population, single individual, and tissue levels. We have further mathematically modelled our data to identify the major drivers determining such heterogeneity. This approach demonstrates that protein translation and degradation rate constants explain the individuality of the heat-shock time-course dynamic. We observed a decline of protein turnover capacity in early adulthood, co-incidentally occurring as the predicted proteostasis collapse. We identified a decline of intestinal response as the tissue that underlies the individual heterogeneity. Additionally, we verified that individuals with enhanced translation fidelity in early adulthood live longer. Altogether, our results reveal that the stochastic onset of proteostasis collapse of somatic tissues during early adulthood reflects individual protein translation capacity underlying heterogenic ageing of isogenic C. elegans.

bioengineering↗

bioRxiv2019

DeltaNeTS+: Elucidating the mechanism of drugs and diseases using gene expression and transcriptional regulatory networks

BackgroundKnowledge on the molecular targets of diseases and drugs is crucial for elucidating disease pathogenesis and mechanism of action of drugs, and for driving drug discovery and treatment formulation. In this regard, high-throughput gene transcriptional profiling has become a leading technology, generating whole-genome data on the transcriptional alterations caused by diseases or drug compounds. However, identifying direct gene targets, especially in the background of indirect (downstream) effects, based on differential gene expressions is difficult due to the complexity of gene regulatory network governing the gene transcriptional processes.\n\nResultsIn this work, we developed a network analysis method, called DeltaNeTS+, for inferring direct gene targets of drugs and diseases from gene transcriptional profiles. DeltaNeTS+ relies on a gene regulatory network model to identify direct perturbations to the transcription of genes. Importantly, DeltaNeTS+ is able to combine both steady-state and time-course gene expression profiles, as well as to leverage information on the gene network structure that is increasingly becoming available for a multitude of organisms, including human. We demonstrated the power of DeltaNeTS+ in predicting gene targets using gene expression data in complex organisms, including Caenorhabditis elegans and human cell lines (T-cell and Calu-3). More specifically, in an application to time-course gene expression profiles of influenza A H1N1 (swine flu) and H5N1 (avian flu) infection, DeltaNeTS+ shed light on the key differences of dynamic cellular perturbations caused by the two influenza strains.\n\nConclusionDeltaNeTS+ is an enabling tool to infer gene transcriptional perturbations caused by diseases and drugs from gene transcriptional profiles. By incorporating available information on gene network structure, DeltaNeTS+ produces accurate predictions of direct gene targets from a small sample size (~10s). DeltaNeTS+ can freely downloaded from http://www.github.com/cabsel/deltanetsplus.

bioinformatics↗

bioRxiv2019

Metabolic stress is a primary pathogenic event in transgenic Caenorhabditis elegans expressing neuronal human amyloid-β

Alzheimers disease (AD) is the most common neurodegenerative disease affecting the elderly worldwide. Mitochondrial dysfunction has been proposed as a key event in the etiology of AD. We have previously modeled amyloid-beta (A{beta})-induced mitochondrial dysfunction in a transgenic Caenorhabditis elegans strain by expressing human A{beta} peptide specifically in neurons (GRU102). Here, we focus on a deeper analysis of these metabolic changes associated with A{beta}-induced mitochondrial dysfunction. Integrating metabolomics, transcriptomics, biochemical studies and computational modeling, we identify alterations in Tricarboxylic Acid (TCA) cycle metabolism following even low-level A{beta} expression. In particular, GRU102 show reduced activity of a rate-limiting TCA cycle enzyme, alpha-ketoglutarate dehydrogenase. These defects are associated with elevation of protein carbonyl content specifically in mitochondria. Importantly, metabolic failure occurs before any significant increase in global protein aggregate is detectable. Treatment with an antidiabetes drug, Metformin, reverses A{beta}-induced metabolic defects, reduces protein aggregation and normalizes the lifespan of GRU102. Our results point to metabolic dysfunction as an early and causative event in AD pathology and a promising target for intervention.

neuroscience↗

bioRxiv2018

Clonal expansion of mitochondrial DNA deletions is a private mechanism of ageing in long-lived animals

Disruption of mitochondrial metabolism and loss of mitochondrial DNA (mtDNA) integrity are widely considered as evolutionarily conserved (public) mechanisms of ageing (Lopez-Otin et al. 2013). Human ageing is associated with loss in skeletal muscle mass and function (Sarcopenia), contributing significantly to morbidity and mortality. Muscle ageing is associated with loss of mtDNA integrity. In humans, clonally expanded mtDNA deletions co-localize with sites of fiber-breakage and atrophy in skeletal muscle. mtDNA deletions may therefore play an important, possibly causal role in sarcopenia. The nematode Caenorhabditis elegans also exhibits age-dependent decline in mitochondrial function and a form of sarcopenia. However, it is unclear if mtDNA deletions play a role in C. elegans ageing. Here we report identification of 266 novel mtDNA deletions in ageing nematodes. Analysis of the mtDNA mutation spectrum and quantification of mutation burden indicates that (1) mtDNA deletions in nematode is extremely rare, (2) there is no significant age-dependent increase in mtDNA deletions and (3) there is little evidence for clonal expansion driving mtDNA deletion dynamics. Thus, mtDNA deletions are unlikely to drive the age-dependent functional decline commonly observed in C. elegans. Computational modelling of mtDNA dynamics in C. elegans indicates that the lifespan of short-lived animals such as C. elegans is likely too short to allow for significant clonal expansion of mtDNA deletions. Together, these findings suggest that clonal expansion of mtDNA deletions is likely a private mechanism of ageing predominantly relevant in long-lived animals such as humans and rhesus monkey and possibly in rodents.

genetics↗

bioRxiv2018

CALISTA: Clustering And Lineage Inference in Single-Cell Transcriptional Analysis

We present CALISTA (Clustering and Lineage Inference in Single-Cell Transcriptional Analysis), a numerically efficient and highly scalable toolbox for an end-to-end analysis of single-cell transcriptomic profiles. CALISTA includes four essential single-cell analyses for cell differentiation studies, including single-cell clustering, reconstruction of cell lineage specification, transition gene identification, and pseudotemporal cell ordering. In these analyses, we employ a likelihood-based approach where single-cell mRNA counts are described by a probabilistic distribution function associated with stochastic gene transcriptional bursts and random technical dropout events. We evaluated the performance of CALISTA by analyzing single-cell gene expression datasets from in silico simulations and various single-cell transcriptional profiling technologies, comprising a few hundreds to tens of thousands of cells. A comparison with existing single-cell expression analyses, including MONOCLE 2 and SCANPY, demonstrated the superiority of CALISTA in reconstructing cell lineage progression and ordering cells along cell differentiation paths. CALISTA is freely available on https://www.cabselab.com/calista.

bioinformatics↗