bioRxiv Science⌕ Search

Biology subjects

Wu, Y.

Publications and source records attributed to Wu, Y..

At least 451 records · Page 25Linked to original sources

Simplified and Rapid Workflow Enhances Throughput of De Novo Sequencing of COVID-19 Neutralizing Antibodies

During the current COVID-19 pandemic, precise antibody sequencing is crucial for the rapid development of broad-spectrum or pan-{beta}-coronavirus neutralizing antibodies (NAbs) to prevent new variants of the coronavirus and other highly pathogenic {beta}-coronaviruses. However, mass spectrometry-based de novo sequencing remains challenging due to its high cost, low throughput, and unexpected missing of certain ion information. We developed an innovative approach using solely bottom-up to provide a rapid, robust, and refined solution for the current de novo sequencing challenges. The methodology, referred to as SP-MEGD for Single-Pot and Multi-Enzymatic Gradient Digestion, capitalizes on a five-protease gradient digestion by sampling every two hours for a total of 6 hours in an integrative reactor. The SP-MEGD method could engender numerous missed cleavage events and produce various peptide products of diverse lengths with overlapping stretches of residues, enabling efficient database-free de novo sequencing. Antibody assembly for single or mixed COVID-19-NAbs with publicly available sequences and commercial antibodies with unknown sequences was efficiently deciphered using our previously proposed optimization method, Fusion assembler. Our innovative study represents the first successful simultaneous discrimination and achieving over 99% accurate sequence coverage of a mixture containing three humanized COVID-19-NAbs (S2P6LH, BD5514LH, and BD5840LH). Furthermore, we utilized SP-MEGD and Fusion to sequence commonly used anti-CD8 and anti-CD4, and successfully confirmed their impact on immune cell clearance through in vivo experiments. Overall, our developed workflow demonstrates promise in facilitating the discovery and development of vaccines and antibody therapeutics while providing deeper insights into monoclonal antibodies (mAbs).

bioengineering↗

Identification of a novel resistance gene which provides insight into Vip3Aa mode of action in Helicoverpa armigera

The global reliance on Bacillus thuringiensis (Bt) proteins for controlling lepidopteran pests in cotton, corn, and soybean crops underscores the critical need to understand resistance mechanisms. Vip3Aa, one of the most widely deployed and currently effective Bt proteins in genetically modified crops, plays a pivotal role in pest management. This study identifies the molecular basis of Vip3Aa resistance in Australian Helicoverpa armigera through genetic crosses, and integrated genomic and transcriptomic analyses. We identified a previously uncharacterized gene, LOC110373801 (designated HaVipR1), as a crucial determinant of Vip3Aa resistance in two field-derived resistant lines. Functional validation using CRISPR-Cas9 knockout in susceptible lines confirmed the genes role in conferring resistance. Despite extensive laboratory selection of Vip3Aa-resistant colonies in Lepidoptera, the biochemical mechanisms underlying resistance have remained elusive. Our research demonstrates that HaVipR1-mediated resistance operates independently of known resistance genes, including midgut-specific chitin synthase and the transcription factor SfMyb. The identification of HaVipR1 offers further insights into the Vip3Aa mechanism of action. This discovery is vital for devising strategies to counteract resistance and sustain the efficacy of Bt crops. Future research should focus on elucidating the biochemical pathways involving HaVipR1 and investigating its interactions with other resistance mechanisms. Our findings underscore the utility of analysing field-derived resistant lines in providing biologically relevant insights and stress the necessity for comprehensive management strategies to maintain agricultural productivity. Significance StatementThis is the first identification of a specific gene in Helicoverpa armigera which mediates resistance to the Bacillus thuringiensis (Bt) protein Vip3Aa. We identify that this gene is disrupted in two different ways in separate field-derived resistant lines, one being a large transposable element insertion in the first intron of the HaVipR1 gene, confirmed using long-read sequencing. Disruption of this gene using CRISPR-Cas9 knockout in a susceptible H. armigera line confers Vip3A resistance. The identification of a specific gene is important for molecular monitoring and management of H. armigera as well as other global pests of concern like Spodoptera frugiperda. These findings also offer insights for researchers aiming to understand how Vip3Aa functions, as the action pathway remains unclear.

genomics↗

Zinc and Metformin Co-Functionalized Polyetheretherketone: A Novel Dental Implant Material Tailored for the Elderly

This study focuses on addressing the challenges of dental implants in the geriatric population by enhancing the bioactivity of polyetheretherketone (PEEK) through surface modification. PEEK, with its elastic modulus close to alveolar bone, mitigates stress shielding but faces limitations in osseointegration due to low bioactivity. We introduced zinc (Zn) and metformin (MF) onto PEEK surfaces via a dopamine-assisted physical adhesion method, creating a functionalized derivative called ZnMF@PEEK. This combination targets diminished osteogenic potential, persistent inflammation, and cell senescence, which are common issues in elderly patients. Comprehensive physicochemical characterizations confirmed the successful preparation of ZnMF@PEEK, and in vitro and in vivo experiments systematically evaluated its biocompatibility and bioactivity. The results indicate that ZnMF@PEEK holds promise as a dental implant material tailored to the specific needs of the elderly, addressing multifaceted challenges in osseointegration.

bioengineering↗

Multi-Modal CLIP-Informed Protein Editing

AO_SCPLOWBSTRACTC_SCPLOWProteins govern most biological functions essential for life, but achieving controllable protein discovery and optimization remains challenging. Recently, machine learning-assisted protein editing (MLPE) has shown promise in accelerating optimization cycles and reducing experimental workloads. However, current methods struggle with the vast combinatorial space of potential protein edits and cannot explicitly conduct protein editing using biotext instructions, limiting their interactivity with human feedback. To fill these gaps, we propose a novel method called ProtET for efficient CLIP-informed protein editing through multi-modality learning. Our approach comprises two stages: in the pretraining stage, contrastive learning aligns protein-biotext representations encoded by two large language models (LLMs), respectively. Subsequently, during the protein editing stage, the fused features from editing instruction texts and original protein sequences serve as the final editing condition for generating target protein sequences. Comprehensive experiments demonstrated the superiority of ProtET in editing proteins to enhance human-expected functionality across multiple attribute domains, including enzyme catalytic activity, protein stability and antibody specific binding ability. And ProtET improves the state-of-the-art results by a large margin, leading to significant stability improvements of 16.67% and 16.90%. This capability positions ProtET to advance real-world artificial protein editing, potentially addressing unmet academic, industrial, and clinical needs.

bioinformatics↗

Transform the Information Landscape: Applying Waddington Landscape to Information Theory of Aging and Epigenetic Rejeuvanation

The Information Theory of Aging (ITOA) provides a framework for understanding aging and rejuvenation via changes in epigenetic information. These changes are visualized using Waddington landscape -- a 3D topology that represents cell trajectories during changes in cell identity -- constructed from field decomposition of scRNA velocity data. Studies have demonstrated that the Waddington landscape can serve as a mechanistic model of epigenetic information. It explains how age-driven decline in cellular function and reprogramming-induced rejuvenation can be understood by studying the transformations in the Waddington landscape. This study provides insights into the representation of epigenetic information and discusses how it can be reprogrammed to restore a younger cellular state.

developmental biology↗

Modulation Of SUDEP By Central Serotonergic Cooperating with Noradrenergic Circuits: A Synergistic-Dependent Manner

Sudden unexpected death in epilepsy (SUDEP) is the leading cause of death in refractory epilepsy patients. Despite previous accumulating evidence has shown that seizure-induced respiratory arrest (S-IRA) may play the main contributor to SUDEP as an initiating event preeminent cause of mortality, the specific underlying mechanism of action remains unclear. Based on our previous work, serotonin (5-HT) signaling in the dorsal raphe nucleus (DRN) is strongly implicated in S-IRA in animal models, including the DBA/1 mice, on the meanwhile, norepinephrine (NE) neurons of the locus coeruleus (LC) also plays a vital role in regulating respiratory function on its own. Superficially, monoaminergic neuron, as important neurotransmitters in the central nervous system, have similar modes of action in the maintenance of nervous system balance, and each of them has a regulatory effect on SUDEP. However, it remains to be investigated whether monoaminergic neuron family (NE and 5-HT) are related in the mechanism of regulating SUDEP, what is even more curious is whether the two are intrinsically linked. Thus, we hypothesize neural mechanism of central noradrenergic and serotonergic circuits in modulating SUDEP in a synergistic-dependent manner, this endeavor will culminate in a significant breakthrough in elucidating the precise mechanism of action underlying SUDEP. In our study, we will use chemogenetics, optogenetics, calcium signal recording, and bidirectional tracing to explore the internal mechanism of DR-LC regulating the occurrence of SUDEP, and by specifically injecting 5-HT2AR antagonist Ketanserin (KET) and/or NE-1R antagonist Prazosin into the pre-Botzinger complex (PBC), it was finally elucidate that the DR-LC-PBC network can effectively reduce the incidence of SIRA. We firstly proposed a powerful target for exploring the reduction of the incidence of SUDEP, which has great clinical translation potential.

neuroscience↗

Identification of targetable vulnerabilities of PLK1-overexpressing cancers by synthetic dosage lethality

Tumor heterogeneity poses a significant challenge in combating treatment resistance. Despite Polo-like kinase 1 (PLK1) being universally overexpressed in cancers and contributing to chromosomal instability (CIN), direct PLK1 inhibition hasnt yielded clinical progress. To address this, we utilized the synthetic dosage lethality (SDL) approach, targeting PLK1s genetic interactions for selective killing of overexpressed tumor cells while mitigating heterogeneity-associated challenges. Employing computational methods, we conducted a genome-wide shRNA screen, identifying 105 SDL candidates. Further in vivo CRISPR screening in a breast cancer xenograft model and in vitro CRISPR analysis validated these candidates. Employing Perturb-seq revealed IGF2BP2/IMP2 as a key SDL hit eliminating PLK1-overexpressing cells. Suppression of IGF2BP2, genetically or pharmacologically, downregulated PLK1 and limited tumor growth. Our findings strongly propose targeting PLK1s genetic interactions as a promising therapeutic approach, holding broad implications across multiple cancers where PLK1 is overexpressed.

cancer biology↗

Identifying Artifacts from Large Library Docking

While large library docking has discovered potent ligands for multiple targets, as the libraries have grown, the very top of the hit-lists can become populated with artifacts that cheat our scoring functions. Though these cheating molecules are rare, they become ever-more dominant with library growth. Here, we investigate rescoring top-ranked molecules from docking screens with orthogonal methods to identify these artifacts, exploring implicit solvent models and absolute binding free energy perturbation (AB-FEP) as cross-filters. In retrospective studies, this approach deprioritized high-ranking non-binders for nine targets while leaving true ligands relatively unaffected. We tested the method prospectively against results from large library docking AmpC {beta}-lactamase. From the very top of the docking hit lists, we prioritized 128 molecules for synthesis and experimental testing, a mixture of 39 molecules that rescoring flagged as likely cheaters and another 89 that were plausible true actives. None of the 39 predicted cheating compounds inhibited AmpC up to 200{micro}M in enzyme assays, while 57% of the 89 plausible true actives did do so, with 19 of them inhibiting the enzyme with apparent Ki values better than 50{micro}M. As our libraries continue to grow, a strategy of catching docking artifacts by rescoring with orthogonal methods may find wide use in the field. Graphical TOC Entry O_FIG O_LINKSMALLFIG WIDTH=166 HEIGHT=200 SRC="FIGDIR/small/603966v1_ufig1.gif" ALT="Figure 1"> View larger version (25K): org.highwire.dtl.DTLVardef@c2f1a5org.highwire.dtl.DTLVardef@86bc58org.highwire.dtl.DTLVardef@1b739c7org.highwire.dtl.DTLVardef@325b93_HPS_FORMAT_FIGEXP M_FIG C_FIG

biophysics↗

BMAP: a comprehensive and reproducible biomedical data analysis platform

In the realm of biomedical research, efficient data analysis and processing are crucial due to the escalating volume and complexity of data generated by research teams. Managing these vast arrays of localized data presents significant challenges, necessitating precise, efficient, and reproducible analysis methodologies to ensure the integrity and reliability of scientific outcomes. Traditional management of analysis codes, computing environments, and the inherent difficulties in result traceability due to team dynamics often lead to inefficiencies and potential risks in maintaining academic integrity. Furthermore, while online storage platforms such as Dryad, GitHub, and Docker facilitate data, code, and environment management, they do not inherently guarantee the reproducibility of results, with issues like data incompleteness, forgotten parameters, or software discrepancies posing additional challenges. To address these critical gaps, we developed a BioMedical data Analysis Platform (BMAP) to offer online and localized categorized management of research assets. BMAP enhances workflow efficiency by transforming complex pipelines into user-friendly web applications, promoting consistency and standardization across team analyses. Its comprehensive web analysis module and seamless integration with data and computing resources support automated result reproducibility and visualization. According to the assessment, 1,692 omics-related figures from 101 recent articles, across 45 visualization types, were tested with BMAP, which could cover 37.8% of the types and 64.3% of the figures. BMAP also enables the sharing and enhancement of research methods through its cloud platform, allowing researchers to utilize the previously developed and validated tools, thereby reducing redundant effort and minimizing analytical discrepancies due to methodological differences.

bioinformatics↗

Ongoing genome doubling promotes evolvability and immune dysregulation in ovarian cancer

Whole-genome doubling (WGD) is a critical driver of tumor development and is linked to drug resistance and metastasis in solid malignancies. Here, we demonstrate that WGD is an ongoing mutational process in tumor evolution. Using single-cell whole-genome sequencing, we measured and modeled how WGD events are distributed across cellular populations within tumors and associated WGD dynamics with properties of genome diversification and phenotypic consequences of innate immunity. We studied WGD evolution in 65 high-grade serous ovarian cancer (HGSOC) tissue samples from 40 patients, yielding 29,481 tumor cell genomes. We found near-ubiquitous evidence of WGD as an ongoing mutational process promoting cell-cell diversity, high rates of chromosomal missegregation, and consequent micronucleation. Using a novel mutation-based WGD timing method, doubleTime, we delineated specific modes by which WGD can drive tumor evolution: (i) unitary evolutionary origin followed by significant diversification, (ii) independent WGD events on a pre-existing background of copy number diversity, and (iii) evolutionarily late clonal expansions of WGD populations. Additionally, through integrated single-cell RNA sequencing and high-resolution immunofluorescence microscopy, we found that inflammatory signaling and cGAS-STING pathway activation result from ongoing chromosomal instability and are restricted to tumors that remain predominantly diploid. This contrasted with predominantly WGD tumors, which exhibited significant quiescent and immunosuppressive phenotypic states. Together, these findings establish WGD as an evolutionarily active mutational process that promotes evolvability and dysregulated immunity in late stage ovarian cancer.

cancer biology↗

Association study of Plasmodium falciparum Acetyl-CoA Synthetase Mutations S868G and V949I with antimalarial drugs

Plasmodium falciparum acetyl-CoA synthetase (PfACAS) protein is an important source of acetyl-CoA. We detected the mutations S868G and V949I in PfACAS by whole-genome sequencing analysis in some recrudescent parasites after antimalarial treatment with artesunate and dihydroartemisinin-piperaquine, suggesting that they may confer drug resistance. Using CRISPR/Cas9 technology, we engineered parasite lines carrying the PfACAS S868G and V949I mutations in two genetic backgrounds and evaluated their susceptibility to antimalarial drugs in vitro. The results demonstrated that PfACAS S868G and V949I mutations alone or in combination were not enough to provide resistance to antimalarial drugs.

microbiology↗

Predicting Enzyme Functions Using Contrastive Learning with Hierarchical Enzyme Structure Information

Enzyme functional annotation is a fundamental challenge in biology, and many computational tools have been developed. Accurate function prediction of enzymes relies heavily on sequence and structural information, providing critical insights into enzyme activity and specificity. However, for less studied proteins or proteins with previously uncharacterized functions or multiple activities, most of these tools cannot accurately predict functional annotations, such as enzyme commission (EC) numbers. At the same time, functional hierarchical information between enzyme species categorized based on EC numbers has not been sufficiently investigated. To address these challenges, we propose a machine learning algorithm named EnzHier, which assigns EC numbers to enzymes with better accuracy and reliability compared to state-of-the-art tools. EnzHier cleverly learns the functional hierarchy of enzymes by optimizing triplet loss, enabling it to annotate understudied enzymes confidently and identify confounding enzymes with two or more EC numbers. By incorporating both sequence and structural information, EnzHier enhances its predictive capabilities. We experimentally demonstrate its excellent performance. We anticipate that this tool will be widely used to predict the function of uncharacterized enzymes, thereby advancing many fields such as drug design and discovery and medical diagnostics.

bioinformatics↗

Fluorogenic Aptamer Optimizations on a Massively Parallel Sequencing Platform

Fluorogenic aptamers (FAPs) have become an increasingly important tool in cellular sensing and pathogen diagnostics. However, fine-tuning FAPs for enhanced performance remains challenging even with the structural details provided by X-ray crystallography. Here we present a novel approach to optimize a DNA-based FAP (D-FAP), Lettuce, on repurposed Illumina next-generation sequencing (NGS) chips. When substituting its cognate chromophore, DFHBI-1T, with TO1-biotin, Lettuce not only shows a red-shifted emission peak by 53 nm (from 505 to 558 nm), but also a 4-fold bulk fluorescence enhancement. After screening 8,821 Lettuce variants complexed with TO1-biotin, the C14T mutation is found to exhibit an improved apparent dissociated constant ([Formula] vs. 0.82 {micro}M), an increased quantum yield (QY: 0.62 vs. 0.59) and an elongated fluorescence lifetime ({tau}: 6.00 vs. 5.77 ns), giving 45% more ensemble fluorescence than the canonical Lettuce/TO1-biotin complex. Molecular dynamic simulations further indicate that the {pi}-{pi} stacking interaction is key to determining the coordination structure of TO1-biotin in Lettuce. Our screening-and-simulation pipeline can effectively optimize FAPs without any prior structural knowledge of the canonical FAP/chromophore complexes, providing not only improved molecular probes for fluorescence sensing but also insights into aptamer-chromophore interactions.

biophysics↗

Eyes robustly blink to musical beats like tapping

Auditory-motor synchronization with musical rhythm may extend beyond overt body movements like fingers or feet. Through four experiments combining eye-tracking, neurophysiological and structural imaging approaches with 123 young non-musicians, we demonstrated robust synchronization of spontaneous eye blinks with musical beats. The blinking rate consistently aligned with the beat rate across various tempi and independent of melodic cues. Blinking exhibited beat phase-specific tuning, with inhibition immediately before beats and increased frequency afterward. Variations in blink-beat synchronization corresponded to the difference in microstructural lateralization of auditory-parietal connectivity. EEG recordings revealed a dynamic correspondence between blink and neural beat tracking. Mechanistically, this synchronization reflects dynamic auditory attention and temporal modulation of visual sampling. Our findings establish eye tapping as a novel behavioral paradigm, expanding the behavioral repertoire of auditory-motor synchronization. It underscores the intricate relationship between music rhythms and peripheral oculomotor system, proposing a cross-modal active sensing and embodied experience in music perception.

neuroscience↗

Characterizing Excretory-Secretory Products Proteome Across Larval Development Stages in Ascaris suum

IntroductionAscaris lumbricoides and Ascaris suum are parasitic nematodes that primarily infest the small intestines of humans and pigs, respectively. Ascariasis poses a significant threat to human health and swine health. Understanding Ascaris larval development is crucial for developing novel therapeutic interventions that will prevent ascariasis in both humans and pigs. This study aimed to characterize the excretory-secretory (ES) proteome of different Ascaris suum larval stages (L3-egg, L3-lung, L3-trachea) to identify potential targets for intervention to prevent Ascaris-induced global morbidity. MethodsStage-specific larvae were isolated, cultured in vitro and ES-product was collected. Third-stage Ascaris larvae (L3) were isolated from embryonated eggs (L3-egg), isolated from the lungs of Balb/c mice infected with Ascaris suum eggs at day 8 post infection (L3-lungs) and isolated from the trachea of Balb/c mice infected with Ascaris suum eggs at day 12 post infection (L3-trachea). ES products were obtained by culturing larvae. Proteomic analysis was conducted using liquid chromatography-tandem mass spectrometry (LC-MS/MS) and bioinformatic tools including MaxQuant, Perseus, and Andromeda, following a detailed protocol available on GitHub. The analysis encompassed peptide identification, scoring, and quantification against an organism-specific database, with subsequent quality control, correlation assessment, and differential abundance determination using the Amica algorithm. ResultsA total of 58 unique proteins were identified in the ES products. Fourteen proteins were common across all stages, while others were stage-specific. Principal component analysis revealed distinct protein profiles for each stage, suggesting qualitatively different proteomes. Gene ontology analysis indicated stage-specific GO enrichment of specific protein classes, such as nuclear proteins in L3-egg ES products and metabolic enzymes in L3-lung and L3-trachea ES products. DiscussionThis study revealed stage-specific differences in the composition of Ascaris ES products. Further investigation into the functional roles of these proteins and their interactions with host cells is crucial for developing novel therapeutic and diagnostic strategies against ascariasis.

microbiology↗

Phospholipase Cgamma regulates lacrimal gland branching by competing with PI3K in phosphoinositide metabolism

Although the regulation of branching morphogenesis by spatially distributed cues is well established, the role of intracellular signaling in determining the branching pattern remains poorly understood. In this study, we investigated the regulation and function of phospholipase C gamma (PLC{gamma}) in Fibroblast Growth Factor (FGF) signaling in lacrimal gland development. We showed that deletion of PLC{gamma}1 in the lacrimal gland epithelium leads to ectopic branching and acinar hyperplasia, which was phenocopied by either mutating the PLC{gamma}1 binding site on Fgfr2 or disabling any of its SH2 domains. PLC{gamma}1 inactivation did not change the level of Fgfr2 or affect MAPK signaling, but instead led to sustained AKT phosphorylation due to increased PIP3 production. Consistent with this, PLC{gamma}1 mutant phenotype can be reproduced by elevation of PI3K signaling in Pten knockout and attenuated by blocking AKT signaling. This study demonstrated that PLC{gamma} modulates PI3K signaling by shifting phosphoinositide metabolism, revealing an important role of signaling dynamics in conjunction with spatial cues in shaping branching morphogenesis.

developmental biology↗

Crk mediates Csk-Hippo signaling independently of Yap tyrosine phosphorylation to induce cell extrusion

Src family kinases (SFKs), including Src, Fyn and Yes, play important roles in development and cancer. Despite being first discovered as the Yes-associated protein, the regulation of Yap by SFKs remains poorly understood. Here, through single-cell analysis and genetic lineage tracing, we show that the pan-epithelial ablation of C-terminal Src kinase (Csk) in the lacrimal gland unleashes broad Src signaling but specifically causes extrusion and apoptosis of acinar progenitors at a time when they are shielded by myoepithelial cells from the basement membrane. Csk mutants can be phenocopied by constitutively active Yap and rescued by deleting Yap or Taz, indicating a significant functional overlap between Src and Yap signaling. Although Src-induced tyrosine phosphorylation has long been believed to regulate Yap activity, we find that mutating these tyrosine residues in both Yap and Taz fails to perturb mouse development or alleviate the Csk lacrimal gland phenotype. In contrast, Yap loses Hippo signaling-dependent serine phosphorylation and translocates into the nucleus in Csk mutants. Further chemical genetics studies demonstrate that acute inhibition of Csk enhances Crk/CrkL phosphorylation and Rac1 activity, whereas removing Crk/CrkL or Rac1/Rap1 ameliorates the Csk mutant phenotype. These results show that Src controls Hippo-Yap signaling through the Crk/CrkL-Rac/Rap axis to promote cell extrusion.

developmental biology↗

Deep Learning-Based Classification of CRISPR Loci Using Repeat Sequences

With the widespread application of the CRISPR-Cas system in gene editing and related fields, the demand for detecting and classifying CRISPR-Cas systems in metagenomic data has continuously increased. The traditional classification of the CRISPR-Cas system mainly relies on identifying neighboring cas genes of repeats. However, in some cases where there is a lack of information about cas genes, such as in metagenomes and fragmented genome assemblies, traditional classification methods may become ineffective. Here, we introduce a deep learning-based method called CRISPRclassify-CNN-Att, which classifies CRISPR-Cas systems solely based on repeat sequences. CRISPRclassify-CNN-Att utilizes convolutional neural networks (CNNs) and self-attention mechanisms to extract features from repeat sequences. It employs a stacking strategy to handle sample imbalances across different subtypes and improves classification accuracy for subtypes with fewer samples through transfer learning. CRISPRclassify-CNN-Att demonstrates excellent performance in classifying multiple subtypes, particularly in subtypes with a larger number of samples. Although CRISPR loci classification primarily relies on cas genes, CRISPRclassify-CNN-Att offers a new approach as a significant complement to current methods. It can identify unclassified loci missed by traditional cas-based methods, breaking the limitations of traditional approaches, and simplifying the classification process. The proposed tool is freely accessible via https://github.com/Xingyu-Liao/CRISPRclassify-CNN-Att.

bioinformatics↗