Biology subjects

Parmar, J.

Publications and source records attributed to Parmar, J..

3 recordsLinked to original sources

bioRxiv2024

scaLR: a low-resource deep neural network-based platform for single cell analysis and biomarker discovery

PurposeSingle-cell RNA sequencing (scRNA-seq) is producing vast amounts of individual cell profiling data. Analysis of such datasets presents a significant challenge in accurately annotating cell types and their associated biomarkers. scRNA-seq datasets analysis will help us understand diseases such as Alzheimers, Cancer, Diabetes, Coronavirus disease 2019 (COVID-19), Systemic Lupus Ery-thematosus (SLE), etc. Recently different pipelines based on machine learning (ML) and Deep Neural Network (DNN) methods have been employed to tackle these issues utilizing scRNA-seq datasets. These pipelines have arisen as a promising resource and are capable of extracting meaningful and concise features from noisy, diverse, and high-dimensional data to enhance annotations and subsequent analysis. Existing tools require high computational resources to execute large sample datasets. MethodsWe have developed a cutting-edge platform known as scaLR (Single Cell Analysis using Low Resource) that efficiently processes data in batches, and reduces the required resources for processing large datasets and running NN models. scaLR is equipped with data processing, feature extraction, training, evaluation, and downstream analysis. The data processing module consists of sample-wise & standard scaler normalization and splitting of data. Its novel feature extraction algorithm, first trains the model on a feature subset and stores feature importance for all the features in that subset. At the end of this process, top K features are selected based on their importance. The model is trained on top K features, its performance evaluation and associated downstream analysis provide significant biomarkers for different cell types and diseases/traits. ResultsTo showcase the capabilities of scaLR, we utilized several scRNA-seq datasets of Peripheral Blood Mononuclear Cells (PBMCs), Alzheimers patients, and large datasets from human and mouse embryonic development. Our findings indicate that scaLR offers comparable prediction accuracy and requires less model training time and compute resources than existing Python-based pipelines and frameworks. Moreover, scaLR efficiently handles large sample datasets (>11.4 million cells) with minimal resource usage (29GB RAM, 12GB GPU, and 8 CPUs) while maintaining high prediction accuracy and being capable of ranking the biomarker association with specific cell types and diseases. ConclusionWe present scaLR a Python-based platform, engineered to utilize minimal computational resources while maintaining comparable execution times to existing frameworks. It is highly scalable and capable of efficiently handling datasets containing millions of cell samples and providing their classification and important biomarkers.

bioinformatics↗

bioRxiv2021

SMC protein RecN drives translocation and remodelling of RecA filament for homology search

While the molecular repertoire of the homologous recombination pathway is well-studied, the search mechanism that enables recombination between distant homologous regions is poorly understood. Earlier work suggests that the recombinase RecA, an essential component for homology search, forms an elongated filament, nucleating at the break site. How this RecA structure carries out long distance search remains unclear. Here, we follow the dynamics of RecA after induction of a single double-strand break on the Caulobacter chromosome. We find that the RecA-nucleoprotein filament, once formed, rapidly translocates in a directional manner in the cell, undergoing several pole-to-pole traversals, until homology search is complete. Concomitant with translocation, we observe dynamic remodelling of the filament. Importantly in vivo, the RecA filament alone is incapable of such long distance movement; both translocation and dynamic remodelling are contingent on action of SMC-like protein RecN, via its ATPase cycle. We provide a stochastic description of RecN-regulated changes in filament length during translocation via modulation of RecA assembly-disassembly. In summary, we have uncovered the three key elements of homology search driven by RecN: mobility of a finite segment of RecA, filament remodelling and ability to conduct multiple pole-to-pole traversals, which together point to a novel optimal search strategy.

cell biology↗

bioRxiv2020

Clustering and reverse transcription of HIV-1 genomes in nuclear niches of macrophages

In order to replicate, the Human Immunodeficiency Virus (HIV-1) reverse transcribes its RNA genome into DNA, which subsequently integrates into host cell chromosomes. These two key events of the viral life cycle are commonly viewed as separate not only in time but also in cellular space, since reverse transcription (RT) is thought to be completed in the cytoplasm before nuclear import and integration. However, the spatiotemporal organization of the early replication cycle in macrophages, natural non-dividing target cells that constitute reservoirs of HIV-1 and an obstacle to curing AIDS, remains unclear. Here, we demonstrate that infected macrophages display large nuclear foci of viral DNA and viral RNA, in which multiple genomes cluster together. These clusters form in the absence of chromosomal integration, sequester the paraspeckle protein CPSF6 and localize to nuclear speckles. Strikingly, we show that viral RNA clusters consist mostly of genomic, incoming RNA, both in cells where RT is pharmacologically suppressed and in untreated cells. We demonstrate that, after temporary inhibition, RT can resume in the nucleus and lead to vDNA accumulation in these clusters. We further show that nuclear RT can result in transcription competent viral DNA. These findings change our understanding of the early HIV-1 replication cycle, and may have implications for understanding HIV-1 persistence.

microbiology↗