Biology subjects

Kandel, S.

Publications and source records attributed to Kandel, S..

7 recordsLinked to original sources

bioRxiv2026

RNA-seq meta-analysis and machine learning identify stress-responsive genes and improve genomic prediction in common bean (Phaseolus vulgaris L.) with cross-species application in cowpea (Vigna unguiculata L.)

Common bean (Phaseolus vulgaris L.) is exposed to a broad spectrum of abiotic and biotic stresses that impose severe constraints on productivity, yet the molecular basis of stress tolerance remains poorly resolved, with independent studies yielding inconsistent and incomplete conclusions. To establish a comprehensive picture of the common bean stress transcriptome, we conducted a systematic meta-analysis of publicly available RNA-sequencing datasets spanning abiotic and biotic stress conditions across leaf and root tissues. Integrating statistical meta-analysis with machine-learning approaches, we identified stress-responsive gene sets whose robustness was verified through rigorous statistical approaches including independent dataset validation. Beyond confirming established stress-responsive genes, the machine-learning framework uncovered candidates overlooked by standard significance thresholds in individual studies yet carrying consistent transcriptional signals across studies. Co-expression and protein-protein network analyses further resolved these candidates into functionally coherent modules linked to specific stress-response programs. Notably, ethylene-responsive transcription factors were identified as hub genes in three of four stress-tissue groups, with NAC domain transcription factors emerging as additional hub genes in biotic stress contexts. Importantly, the biological significance of the identified gene sets was validated genomically: marker panels targeting consensus meta-analysis-derived and machine-learning-discovered gene regions improved genomic prediction accuracy for disease resistance traits in common bean and abiotic stress tolerance traits in cowpea relative to a baseline model with equivalent-sized random marker sets. Overall, these findings revealed conserved stress transcriptome signatures in common bean and provided a cross-species, evidence-based framework for prioritizing candidate genes and constructing biologically informed genomic selection tools to advance stress-resilient legume breeding.

genetics↗

bioRxiv2026

Targeting EZH2 Oncogenic Splicing: Decoding the Regulatory Network and Antisense Correction

Recurrent mutations in splicing factors (SFs) have been established as crucial drivers of tumorigenesis in several types of blood cancer, and also common in a variety of solid tumors. Mutations change the RNA-binding preferences of SFs, promote global splicing alterations, and often generate erroneous mRNAs that are then degraded by nonsense-mediated mRNA decay (NMD). Consequently, several critical genes linked to hematopoiesis are dysregulated, leading to blood cancer. Although the field has progressed considerably in identifying aberrant genes and affected pathways, effective therapies have not yet emerged in SF-mutated cancers. To address this key gap, we instigated a gene-specific targeted strategy by unlocking the regulatory network. As a proof-of-concept, we scrutinized a tumor suppressor gene EZH2, which is a bona fide target in SRSF2-mutated cancer. We precisely defined splicing cis-elements in EZH2 transcripts and illustrated the dynamic choreography of regulatory proteins in the entire splicing and NMD catalytic pathways. We uncovered a highly coordinated cross-regulation between splicing and NMD promoted by mutant SRSF2 by enhancing the deposition of critical spliceosome- and NMD-associated factors, augmenting mRNA decay to ablate tumor suppression. We then designed antisense oligonucleotides (ASOs) targeting important regulatory sites. Our lead ASO successfully corrects aberrant splicing and NMD, restores the expression and function of EZH2, and partially rescues hematopoietic defects and cellular properties. Our study demonstrates that ASO pharmacology is an actionable strategy for clinical development, challenging the existing paradigms in SF-mutated cancers.

cancer biology↗

bioRxiv2025

Lessons learned from a Kaggle challenge for particle picking in cryo-electron tomography

The difficulty of particle picking in cryo-electron tomography remains a barrier to routine in situ structure determination. Machine learning is well-suited to overcome this bottleneck with efficient algorithms that generalize across molecular species. To spur new algorithm development, we held a three-month Kaggle challenge that tasked contestants with annotating five molecular species across hundreds of experimental tomograms. This competition successfully engaged >1000 participants from diverse fields and delivered particle pickers that outperformed existing state-of-the-art. Systematic comparisons of the contestants submissions revealed the tolerance of subtomogram averaging to moderate but not severe over-picking and underscored the need for more robust measures of annotation quality. The winning models also highlighted the importance of data augmentation to overcome limited training data. All competition tomograms along with the ground truth and winning teams annotations have been released on the CryoET Data Portal as a resource to benchmark current and future tools for particle picking.

cell biology↗

bioRxiv2024

Annotating CryoET Volumes: A Machine Learning Challenge

Cryo-electron tomography (cryoET) has emerged as a powerful structural biology tool for understanding protein complexes in their native cellular environments. Presently, 3D volumes of cellular environments can be acquired in the thousands in a few days where each volume provides a rich and complex cellular landscape. Despite numerous innovations, localizing and identifying the vast majority of protein species in these volumes remains prohibitively difficult. Machine learning based methods provide an opportunity to automate the process of labeling and annotating cryoET volumes. Due to current bottlenecks in the annotation process, and a lack of large standardized datasets, training datasets for machine learning algorithms have been scarce. Here, we present a defined "phantom" sample, along with "ground truth" annotations, that will be the basis of a machine learning challenge to bring cryoET and ML experts together and spur creativity to address this annotation problem. We have also set up a cryoET data portal that provides additional diverse sets of annotated 3D volumes from cryoET experts across the world for the machine learning challenge.

cell biology↗

bioRxiv2024

Open-source Tools for CryoET Particle Picking Machine Learning Competitions

We are launching a machine learning (ML) competition focused on particle picking in cryo-electron tomography (cryoET) data, a crucial task in structural biology. To support this, we have created a comprehensive suite of open-source tools to develop resources for our competition, including copick for dataset management, napari plugins for interactive visualization, utilities for converting particle picks to segmentation masks, and PyTorch tools for custom dataset sampling. These resources streamline the processes of data handling, labeling, and visualization, allowing participants to focus on model development. By leveraging these tools, competitors will be better equipped to tackle the unique challenges of cryoET data and push forward advancements in particle picking techniques.

bioinformatics↗

bioRxiv2023

Genomic Surveillance of SARS-CoV-2 Using Long-Range PCR Primers

Whole Genome Sequencing (WGS) of the SARS-CoV-2 virus is crucial in the surveillance of the COVID-19 pandemic. Several primer schemes have been developed to sequence the [~]30,000 nucleotide SARS-CoV-2 genome that use a multiplex PCR approach to amplify cDNA copies of the viral genomic RNA. Midnight primers and ARTIC V4.1 primers are the most popular primer schemes that can amplify segments of SARS-CoV-2 (400 bp and 1200 bp, respectively) tiled across the viral RNA genome. Mutations within primer binding sites and primer-primer interactions can result in amplicon dropouts and coverage bias, yielding low-quality genomes with Ns inserted in the missing amplicon regions, causing inaccurate lineage assignments, and making it challenging to monitor lineage-specific mutations in Variants of Concern (VoCs). This study uses seven long-range PCR primers with an amplicon size of [~]4500 bp to tile across the complete SARS-CoV-2 genome. One of these regions includes the full-length S-gene by using a set of flanking primers. Using a small set of long-range primers to sequence SARS-CoV-2 genomes reduces the possibility of amplicon dropout and coverage bias.

genomics↗

bioRxiv2018

Rapid Sequencing of Multiple RNA Viruses in their Native Form

Long-read nanopore sequencing by a MinION device offers the unique possibility to directly sequence native RNA. We combined an enzymatic poly-A tailing reaction with the native RNA sequencing to (i) sequence complex population of single-stranded (ss)RNA viruses in parallel, (ii) detect genome, subgenomic mRNA/mRNA simultaneously, (iii) detect a complex transcriptomic architecture without the need for assembly, (iv) enable real-time detection. Using this protocol, positive-ssRNA, negative-ssRNA, with/without a poly(A)-tail, segmented/non-segmented genomes were mixed and sequenced in parallel. Mapping of the generated sequences on the reference genomes showed 100% length recovery with up to 97 % identity. This work provides a proof of principle and the validity of this strategy, opening up a wide range of applications to study RNA viruses.

genomics↗