Biology subjects

Prasad, G.

Publications and source records attributed to Prasad, G..

15 recordsLinked to original sources

bioRxiv2026

Inferring ecDNA Evolution from Allele Frequency Spectra

Extrachromosomal DNA (ecDNA) can arise early in the development of aggressive tumors or later upon progression, but the impact of chronology on patient outcome is not known. Currently, it is not possible to determine when an ecDNA arose in the course of tumor development and progression, based on genome sequencing data. Here, we develop an analytic method, ecEvol, to infer ecDNA emergence time, selection strength, mutation rate, and clonal dominance from allele frequency spectra in bulk whole-genome sequencing. EcEvol outperforms other methods and predicts parameters with high accuracy across a wide range of parameter values. Applying ecEvol to primary, treatment-naive tumors, we find that ecDNA typically emerges early and undergoes strong positive selection, albeit with substantial variation across cancer types. EcDNAs carrying known oncogenes emerge earlier and reach greater clonal dominance than ecDNAs lacking known oncogenes. In advanced and predominantly treated tumors, ecDNA emerges substantially later than in matched primary tumors, consistent with the generation and selection of new ecDNA during tumor progression and therapy. Longitudinal samples further support continued expansion of ecDNA-bearing lineages. Finally, earlier ecDNA emergence is associated with poorer overall survival. These results provide insight into the chronology of ecDNA formation from a single tumor biopsy taken at any point during the course of the disease, revealing that allele-frequency spectra encode the evolutionary history of ecDNA and provide a quantitative framework for reconstructing its emergence and expansion.

bioinformatics↗

bioRxiv2026

Fast and accurate resolution of ecDNA sequence using Cycle-Extractor

Extrachromosomal DNA (ecDNA) plays a key role in cancer pathology. EcDNAs mediate high oncogene amplification and expression and worse patient outcomes. Accurately determining the structure of these circular molecules is essential for understanding their function, yet reconstructing ecDNA cycles from sequencing data remains challenging. We introduce Cycle-Extractor (CE) for reconstruction. CE accepts a breakpoint graph derived from either short or long read sequencing data as input and extracts a cycle with the maximum length-weighted-copy-number. CE utilizes a mixed-integer linear program (MILP) and a separate traversal procedure, enabling fast optimization and compatibility with free solvers. We evaluated CE against CoRAL (long-read-based quadratic optimization), Decoil (long-reads), and AmpliconArchitect (AA for short reads) on both simulated data and real cancer cell lines. On simulated ecDNA, CE achieves performance comparable to CoRAL across three accuracy metrics and consistently outperforms AA and Decoil. On cancer cell lines, CE produces longer and heavier cycles than AA, and achieves performance similar to CoRAL. Moreover, CE is, on average, 40x faster than CoRAL. These results demonstrate that CE accurately reconstructs ecDNA from both short- and long-read sequencing data, while long-read inputs allow CE to recover more complete and higher-confidence ecDNA structures. CE improved the prediction of many ecDNA structures. On a PC3 ecDNA containing MYC, CE uses ONT data to reconstruct a substantially larger and higher-copy sequence (4.2 Mbp) compared to the short-read-derived reconstruction (690 Kbp). CRISPR-CATCH experiments confirm the presence of a large ecDNA molecule, validating the long-read-based CE reconstruction.

bioinformatics↗

bioRxiv2025

Accurate Prediction of ecDNA in Interphase Cancer Cells using Deep Neural Networks

Oncogene amplification is a key driver of cancer pathogenesis and is often mediated by extrachromosomal DNA (ecDNA). EcDNA amplifications are associated with increased pathogenicity of cancer and poorer outcomes for patients. EcDNA can be detected accurately using fluorescence in situ hybridization (FISH) when cells are arrested in metaphase. However, the majority of cancer cells are non-mitotic and must be analyzed in interphase, where it is difficult to discern extrachromosomal amplifications from chromosomal amplifications. Thus, there is a need for methods that accurately predict oncogene amplification status from interphase cells. Here, we present interSeg, a deep learning-based tool to cytogenetically determine the amplification status as EC-amp, HSR-amp, or not amplified from interphase FISH images. We trained and validated interSeg on 652 images (40,446 nuclei). Tests on 215 cultured cell and tissue model images (9,733 nuclei) showed 89% and 97% accuracy at the nuclear and sample levels, respectively. The neuroblastoma patient tissue hold-out set (67 samples and 1,937 nuclei) also revealed 97% accuracy at the sample level in detecting the presence of focal amplification. In experimentally and computationally mixed images, interSeg accurately predicted the level of heterogeneity. The results showcase interSeg as an important method for analyzing oncogene amplifications.

cancer biology↗

bioRxiv2024

Predictive Prioritization of Enhancers Associated with Pancreas Disease Risk

Genetic and epigenetic variations in regulatory enhancer elements increase susceptibility to a range of pathologies. Despite recent advances, linking enhancer elements to target genes and predicting transcriptional outcomes of enhancer dysfunction remain significant challenges. Using 3D chromatin conformation assays, we generated an extensive enhancer interaction dataset for the human pancreas, encompassing more than 20 donors and five major cell types, including both exocrine and endocrine compartments. We employed a network approach to parse chromatin interactions into enhancer-promoter tree models, facilitating a quantitative, genome-wide analysis of enhancer connectivity. With these tree models, we developed a machine learning algorithm to estimate the impact of enhancer perturbations on cell type-specific gene expression in the human pancreas. Orthogonal to our computational approach, we perturbed enhancer function in primary human pancreas cells using CRISPR interference and quantified the effects at the single-cell level through RNA FISH coupled with high-throughput imaging. Our enhancer tree models enabled the annotation of common germline risk variants associated with pancreas diseases, linking them to putative target genes in specific cell types. For pancreatic ductal adenocarcinoma, we found a stronger enrichment of disease susceptibility variants within acinar cell regulatory elements, despite ductal cells historically being assumed as the primary cell-of-origin. Our integrative approach--combining cell type-specific enhancer-promoter interaction mapping, computational models, and single-cell enhancer perturbation assays--produced a robust resource for studying the genetic basis of pancreas disorders.

genomics↗

bioRxiv2024

AmpliconSuite: an end-to-end workflow for analyzing focal amplifications in cancer genomes

Focal amplifications in the cancer genome, particularly extrachromosomal DNA (ecDNA) amplifications, are emerging as a pivotal event in cancer progression across diverse cancer contexts, presenting a paradigm shift in our understanding of tumor dynamics. Simultaneously, identification of the various modes of focal amplifications is bioinformatically challenging. We present AmpliconSuite, a collection of tools that enables robust identification of focal amplifications from whole-genome sequencing data. AmpliconSuite includes AmpliconSuite- pipeline; utilizing the AmpliconArchitect (AA) method, and AmpliconRepository.org; a community- editable website for the sharing of focal amplification calls. We also describe improvements made to AA since its initial release that improve its accuracy and speed. As a proof of principle, we utilized publicly available pan-cancer datasets encompassing 2,525 tumor samples hosted on AmpliconRepository.org to illustrate important properties of focal amplifications, showing ecDNA has higher copy number, and stronger oncogene enrichment, compared to other classes of focal amplifications. Finally, we illustrate how AmpliconSuite-pipeline enables delineation of the various mechanisms by which ecDNA forms.

bioinformatics↗

bioRxiv2023

Breakage fusion bridge cycles drive high oncogene copy number, but not intratumoral genetic heterogeneity or rapid cancer genome change

Oncogene amplification is a major driver of cancer pathogenesis. Breakage fusion bridge (BFB) cycles, like extrachromosomal DNA (ecDNA), can lead to high copy numbers of oncogenes, but their impact on intratumoral heterogeneity, treatment response, and patient survival are not well understood due to difficulty in detecting them by DNA sequencing. We describe a novel algorithm that detects and reconstructs BFB amplifications using optical genome maps (OGMs), called OM2BFB. OM2BFB showed high precision (>93%) and recall (92%) in detecting BFB amplifications in cancer cell lines, PDX models and primary tumors. OM-based comparisons demonstrated that short-read BFB detection using our AmpliconSuite (AS) toolkit also achieved high precision, albeit with reduced sensitivity. We detected 371 BFB events using whole genome sequences from 2,557 primary tumors and cancer lines. BFB amplifications were preferentially found in cervical, head and neck, lung, and esophageal cancers, but rarely in brain cancers. BFB amplified genes show lower variance of gene expression, with fewer options for regulatory rewiring relative to ecDNA amplified genes. BFB positive (BFB (+)) tumors showed reduced heterogeneity of amplicon structures, and delayed onset of resistance, relative to ecDNA(+) tumors. EcDNA and BFB amplifications represent contrasting mechanisms to increase the copy numbers of oncogene with markedly different characteristics that suggest different routes for intervention.

cancer biology↗

bioRxiv2023

Identifying the core genome of the nucleus-forming bacteriophage family and characterization of Erwinia phage RAY

We recently discovered that some bacteriophages establish a nucleus-like replication compartment (phage nucleus), but the core genes that define nucleus-based phage replication and their phylogenetic distribution were unknown. By studying phages that encode the major phage nucleus protein chimallin, including previously sequenced yet uncharacterized phages, we discovered that chimallin-encoding phages share a set of 72 highly conserved genes encoded within seven distinct gene blocks. Of these, 21 core genes are unique to this group, and all but one of these unique genes encode proteins of unknown function. We propose that phages with this core genome comprise a novel viral family we term Chimalliviridae. Fluorescence microscopy and cryo-electron tomography studies of Erwinia phage vB_EamM_RAY confirm that many of the key steps of nucleus-based replication encoded in the core genome are conserved among diverse chimalliviruses, and reveal that non-core components can confer intriguing variations on this replication mechanism. For instance, unlike previously studied nucleus-forming phages, RAY doesnt degrade the host genome, and its PhuZ homolog appears to form a five-stranded filament with a lumen. This work expands our understanding of phage nucleus and PhuZ spindle diversity and function, providing a roadmap for identifying key mechanisms underlying nucleus-based phage replication.

microbiology↗

bioRxiv2022

Imaging cortical engagement during motor imagery, mental arithmetic, and silent word generation using MEG beta rhythm

Accurate mapping of cortical engagement during mental imagery or cognitive tasks remains a challenging brain-imaging problem with immediate relevance to the development of brain-computer interfaces (BCI). We analyzed data from fourteen individuals who performed cued motor imagery, mental arithmetic, or silent word generation tasks during MEG recordings. During the motor imagery task, participants imagined the movement of either both hands (HANDS) or both feet (FEET) after the appearance of a static visual cue. During the cognitive task, participants either mentally subtracted two numbers that were displayed on the screen (SUB) or generated words starting with a letter cue that was presented (WORD). The MEG recordings were denoised using a combination of spatiotemporal filtering, the elimination of noisy epochs, and ICA decomposition. Cortical source power in the beta-band (17-25 Hz) was estimated from the selected temporal windows using a frequency-resolved beamforming method applied to the sensor-level MEG signals. The task-related cortical engagement was inferred from beta power decrements within non-overlapping 400 ms temporal windows between 400 and 2800 ms after cue presentation relative to a baseline 400 ms temporal window before cue onset. We estimated the significance of these power changes within each parcel of the Desikan-Killiany atlas using a non-parametric permutation test at the group level. During the HANDS and FEET movement-imagery conditions, beta-power decreased in premotor and motor areas, consistent with a robust engagement of these cortical regions during motor imagery. During WORD and SUB tasks, beta-power decrements signaling cortical engagement were lateralized to left hemispheric brain areas that are expected to engage in language and arithmetic processing within the temporal (superior temporal gyrus), parietal (supramarginal gyrus), and (inferior) frontal regions. A leave-one-subject-out cross-validation using a support vector machine (SVM) applied to beta power decrements across brain parcels yielded accuracy rates of 74% and 68% respectively, for classifying motor-imagery (HANDS-vs-FEET) and cognitive (WORD-vs-SUB) tasks. From the motor-versus-nonmotor contrasts, accuracy rates of 85% and 80% respectively, were observed for HAND-vs-WORD and HAND-vs-SUB. A multivariate Gaussian process classifier (GPC) provided an accuracy rate of 60% for a four-way (HANDS-FEET-WORD-SUB) classification problem. The regions identified by both SVM and GPC classification weight maps were largely consistent with the source modeling findings. Within-subject correlations of beta-decrements during the two task sessions provided insights into the level of engagement by individual subjects and showed moderately high correlations for most subjects. Our results show that it is possible to map the dynamics of cortical engagement during mental processes in the absence of dynamic sensory stimuli or overt behavioral outputs using task-related beta-power decrements. The ability to do so with the high spatiotemporal resolution afforded by MEG could potentially help better characterize the physiological basis of motor or cognitive impairments in neurological disorders and guide strategies for neurorehabilitation.

neuroscience↗

bioRxiv2021

Network analysis after epigenome-wide methylation study revealed JUP as a regulator of co-methylated risk-module for T2DM

Controlling the global Type 2 diabetes mellitus (T2DM) pandemic requires a better understanding of its risk factors across different populations, and needs markers that can precisely predict the individual risk to the disease. DNA methylation due to environmental factors is a known mechanism for conferring risk to T2DM. To identify such methylation signatures and associated risk to the disease, we performed an epigenome-wide-association study (EWAS) in 844 individuals of Indo-European origin. Within the Indian population, we identified and validated 6 novel differentially methylated CpG sites in PDCD6IP, MIR1287, 5S_rRNA, HDAC9, KCNK16, and RTN1 genes associated with T2DM risk at the epigenome-wide-significance-level (p<1.2x10-7). Further, we also replicated the association of 3 known CpG sites in TXNIP, SREBF1 and CPT1A in the Indian population. Using methylation-based-network analysis, we identified 4 co-methylated modules, which we named as turquoise, yellow, brown, and blue, among differentially methylated CpG sites in discovery phase samples. We observed that methylation of the brown module with 28 CpG sites, associated with T2DM risk factors (e.g., BMI, insulin, C-peptide). Upon further analysis, we noted that these methylation signatures at 14 of the brown modules CpG can be used as marker to segregate T2DM patients with good glycemic control (e.g., low HbA1c) but poor lipid profile (low HDL and high TG) from the other patients. Additionally, we discovered that rs6503650 in the JUP gene regulates methylation at all the 14 hub CpG sites of the brown modules as methQTL. Our network-assisted epigenome-wide association study is the first to systematically explore DNA methylations conferring risks to T2DM. In addition, the study shows the potential use of identified risk CpG sites for patient segregation with different clinical outcomes. These findings can be useful for better stratification of patient in order to improve the clinical management and treatment outcomes.

genomics↗

bioRxiv2021

Functional Differences in the Neural Substrates of Auditory Cognition as a Consequence of Music Training

Previous studies have demonstrated that musical deviants (syntactically irregular chords) elicit event related potentials/fields with negative polarity; specifically, the early right anterior negativity and the right anterior temporal negativity responses with peak latencies at ~200 ms and ~350 ms, respectively, post stimulus onset. Here, we investigated differences in the neural dynamics of the auditory perceptual system of individuals with music training compared to those with no music training. Magnetoencephalography was used to examine the neural response to a deviant sound when the auditory system was primed using stimulus entrainment to evoke an auditory gamma-band response between 31 Hz and 39 Hz, in 2 Hz steps. Participants responded to the harmonic relationship between the entrainment stimulus and the subsequent target stimulus. Gamma frequencies carry stimulus information; thus, the paradigm primed the auditory system with a known gamma frequency and evaluated any improvement in the brains response to a deviant stimulus. The entrainment stimuli did not elicit an early right anterior negativity response. Furthermore, the source location of the event-related field difference during the later right anterior temporal negativity response time-window varied depending on group and entrainment condition. In support of previous findings from research using this, and a functionally similar visual-priming paradigm, a 7 Hz phase modulation of gamma amplitude was found for non-musicians following 33 Hz stimulus entrainment. Overall, significant effects of gamma entrainment were found more frequently in the non-music brain. By contrast, musicians demonstrated a greater range of interactions with slower brain rhythms, indicative of increased top-down control. Graphical Abstract O_FIG O_LINKSMALLFIG WIDTH=200 HEIGHT=130 SRC="FIGDIR/small/446932v1_ufig1.gif" ALT="Figure 1"> View larger version (49K): org.highwire.dtl.DTLVardef@1097a58org.highwire.dtl.DTLVardef@13f3c74org.highwire.dtl.DTLVardef@99fae1org.highwire.dtl.DTLVardef@125d351_HPS_FORMAT_FIGEXP M_FIG C_FIG ConclusionO_LIMusicians auditory cognition relies more on top-down processes, while non-musicians rely on bottom-up processing, and therefore, their auditory cognition is facilitated by the entrained gamma-band response. C_LIO_LIFirst neural evidence in support of faster reaction-time responses due to an interaction in phase of an entrained gamma-band response of 33 Hz and a slower endogenous theta rhythm (~7 Hz) - consistently reported in previous research. C_LI

neuroscience↗

bioRxiv2021

Temporal Dynamics of Neural Processing of Facial Expressions and Emotions

Emotion processing has been a focus of research in psychology and neuroscience for some decades. While the evoked neural markers in human brain activations in response to different emotions have been reported, the temporal dynamics of emotion processing has received less attention. Differences in processing speeds, that depend on emotion type, have not been determined. Furthermore, behavioral studies have found that the right side of the human face expresses emotions more accurately than the left side. Therefore, accounting for both the content of the emotion and the visual angle of presentation from the perspective of the viewer, here we have investigated variability in the discrimination of happy and sad faces when the visual angle of presentation was Positive (right side of the face) compared to Negative (left side of the face). Furthermore, the temporal dynamics involved in processing happy and sad emotions have been compared. Regardless of visual angle, happy emotions were processed faster than sad emotions. However, the evoked response to sad emotions significantly increased in amplitude compared to that for happy emotions, when faces were presented at Positive visual angles only. Source reconstruction from sensor-level ERFs show localized activities in ventral and dorsal stream, including fusiform gyrus, lingual gyrus, putamen and Pre and Post central gyrus. Multivariate pattern analysis (MVPA) confirmed these findings - demonstrating successful decoding of happy and sad emotions only occurred when the facial expression was viewed from a positive visual angle, and that happy emotions were processed faster than sad emotions.

neuroscience↗

bioRxiv2019

Cluster-permutation statistical analysis for high-dimensional brain-wide functional connectivity mapping

Brain functional connectivity (FC) analyses based on magnetoencephalographic (MEG) signals have yet to exploit the intrinsic high-dimensional information. Typically, these analyses are constrained to regions of interest to avoid the curse of dimensionality, which leads to conservative hypothesis testing. We removed such constraint by extending cluster-permutation statistics for high-dimensional MEG-FC analyses. We demonstrated the feasibility of this approach by identifying MEG-FC resting-state changes in mild cognitive impairment (MCI), a prodromal stage of Alzheimers disease. We found dense clusters of increased connectivity strength in MCI compared to healthy controls (hypersynchronization), in delta (1-4 Hz) and higher-theta (6-8 Hz) bands oscillations. These clusters mainly consisted of interactions between occipitofrontal and occipitotemporal regions in the left hemisphere and could potentially be used as neuromarkers of early progression in Alzheimers disease. Our novel approach can be used to generate high-resolution statistical FC maps for neuroimaging studies in general.

neuroscience↗

bioRxiv2019

Multifaceted genome-wide study identifies novel regulatory loci for body mass index in Indians

Obesity, a risk factor for various human diseases originates through complex interactions between genes and prevailing environment that varies across populations. Indians exhibit a unique obesity phenotype likely attributed by specific gene pool and environmental factors. Here, we present genome-wide association study (GWAS) of 7,259 Indians to understand the genetic architecture of body mass index (BMI) in the population. Our study revealed novel association of variants in BAI3 (rs6913677) and SLC22A11 (rs2078267) at GWAS significance, and of ZNF45 (rs8100011) with near GWAS significance. As genetic loci may dictate the phenotype through modulation of epigenetic processes, we overlapped discovered genetic signatures with DNA methylation patterns of 236 Indian individuals, and analyzed expression of the candidate genes using publicly available data. The variants in BAI3 and SLC22A11 were found to dictate methylation patterns at unique CpGs harboring critical cis- regulatory elements. Further, BAI3, SLC22A11 and ZNF45 variants were found to overlie repressive chromatin, active enhancer, and active chromatin regions, in that order, in human subcutaneous adipose tissue in ENCODE database. Besides, the identified genomic regions represented potential binding sites for key transcription factors implicated in obesity and/or metabolic disorders. Interestingly, rs8100011 (ZNF45) acted as a robust cis-expression quantitative trait locus (cis-eQTL) in subcutaneous adipose tissue in GTEx portal, and ZNF45 gene expression showed an inverse correlation with BMI in skeletal muscle of Indian subjects. Further, gene-based GWAS analysis revealed CPS1 and UPP2 as additional leads regulating BMI in Indians. Our study decodes potential genomic mechanisms underlying obesity phenotype in Indians.

genomics↗

bioRxiv2019

A practical computerized decision support system for predicting the severity of Alzheimer’s disease of an individual

Computerized clinical decision support systems can help to provide objective, standardized, and timely dementia diagnosis. However, current computerized systems are mainly based on the group analysis, discrete classification of disease stages, or expensive and not readily accessible biomarkers, while current clinical practice relies relatively heavily on cognitive and functional assessments (CFA). In this study, we developed a computational framework using a suite of machine learning tools for identifying key markers in predicting the severity of Alzheimers disease (AD) from a large set of biological and clinical measures. Six machine learning approaches, namely Kernel Ridge Regression (KRR), Support Vector Regression (SVR), and k-Nearest Neighbor (kNNreg) for regression and Support Vector Machine (SVM), Random Forest (RF), and k-Nearest Neighbor (kNNclass) for classification, were used for the development of predictive models. We demonstrated high predictive power of CFA. Predictive performance of models incorporating CFA was shown to be consistently higher accuracy than those based solely on biomarker modalities. We found that KRR and SVM were the best performing regression and classification methods respectively. The optimal SVM performance was observed for a set of four CFA test scores (FAQ, ADAS13, MoCA, MMSE) with multi-class classification accuracy of 83.0%, 95%CI = (72.1%, 93.8%) while the best performance of the KRR model was reported with combined CFA and MRI neuroimaging data, i.e., R2 = 0.874, 95%CI = (0.827, 0.922). Given the high predictive power of CFA and their widespread use in clinical practice, we then designed a data-driven and self-adaptive computerized clinical decision support system (CDSS) prototype for evaluating the severity of AD of an individual on a continuous spectrum. The system implemented an automated computational approach for data pre-processing, modelling, and validation and used exclusively the scores of selected cognitive measures as data entries. Taken together, we have developed an objective and practical CDSS to aid AD diagnosis.

neuroscience↗

bioRxiv2018

Neural Circuit Mechanism of Decision Uncertainty and Change-of-Mind

Decision-making is often accompanied by a degree of confidence on whether a choice is correct. Decision uncertainty, or lack in confidence, may lead to change-of-mind. Studies have identified the behavioural characteristics associated with decision confidence or change-of-mind, and their neural correlates. Although several theoretical accounts have been proposed, there is no neural model that can compute decision uncertainty and explain its effects on change-of-mind. We propose a neuronal circuit model that computes decision uncertainty while accounting for a variety of behavioural and neural data of decision confidence and change-of-mind, including testable model predictions. Our theoretical analysis suggests that change-of-mind occurs due to the presence of a transient uncertainty-induced choice-neutral stable steady state and noisy fluctuation within the neuronal network. Our distributed network model indicates that the neural basis of change-of-mind is more distinctively identified in motor-based neurons. Overall, our model provides a framework that unifies decision confidence and change-of-mind.

neuroscience↗