A5032083590- CRISPR and Genetic Engineering
- RNA and protein synthesis mechanisms
- Cancer Immunotherapy and Biomarkers
- Immunotherapy and Immune Responses
- Advanced biosensing and bioanalysis techniques
- Cancer Genomics and Diagnostics
- Single-cell and spatial transcriptomics
- Computational Drug Discovery Methods
- Machine Learning in Materials Science
- Monoclonal and Polyclonal Antibodies Research
- Cell Adhesion Molecules Research
- RNA regulation and disease
- Innovation and Socioeconomic Development
- vaccines and immunoinformatics approaches
- RNA Interference and Gene Delivery
- AI in cancer detection
- Genetics, Aging, and Longevity in Model Organisms
- Statistical Methods in Clinical Trials
- Bioinformatics and Genomic Networks
- Genomics and Chromatin Dynamics
- Cell Image Analysis Techniques
- Advanced Graph Neural Networks
- Radiomics and Machine Learning in Medical Imaging
- Privacy-Preserving Technologies in Data
- Plant Virus Research Studies
Tongji University
2016-2025
Tongji Hospital
2023-2025
Shanghai East Hospital
2020-2024
Shanghai Tenth People's Hospital
2016-2021
A major challenge for effective application of CRISPR systems is to accurately predict the single guide RNA (sgRNA) on-target knockout efficacy and off-target profile, which would facilitate optimized design sgRNAs with high sensitivity specificity. Here we present DeepCRISPR, a comprehensive computational platform unify sgRNA site prediction into one framework deep learning, surpassing available state-of-the-art in silico tools. In addition, DeepCRISPR fully automates identification...
Abstract Here, we present a multi-modal deep generative model, the single-cell Multi-View Profiler (scMVP), which is designed for handling sequencing data that simultaneously measure gene expression and chromatin accessibility in same cell, including SNARE-seq, sci-CAR, Paired-seq, SHARE-seq, Multiome from 10X Genomics. scMVP generates common latent representations dimensionality reduction, cell clustering, developmental trajectory inference separate imputations differential analysis...
Abstract The powerful CRISPR genome editing system is hindered by its off-target effects, and existing computational tools achieved limited performance in genome-wide prediction due to the lack of deep understanding molecular mechanism. In this study, we propose incorporate dynamics (MD) simulations analysis system, present CRISOT, an integrated tool suite containing four related modules, i.e., CRISOT-FP, CRISOT-Score, CRISOT-Spec, CRISORT-Opti for RNA-DNA interaction fingerprint generation,...
The de novo molecule generation problem involves generating novel or modified molecular structures with desirable properties. Taking advantage of the great representation learning ability deep models, generative which differ from discriminative models in their traditional machine approach, provide possibility molecules directly. Although have been extensively discussed community, a specific investigation computational issues related to for is needed. A concise and insightful discussion...
Abstract Background Cancer neoantigens are expressed only in cancer cells and presented on the tumor cell surface complex with major histocompatibility (MHC) class I proteins for recognition by cytotoxic T cells. Accurate rapid identification of play a pivotal role immunotherapy. Although several silico tools neoantigen prediction have been presented, limitations these exist. Results We developed pTuneos , computational pipeline p rioritizing tu mor neo antigens from next-generation s...
Quantitative structure-activity relationship (QSAR) analysis is commonly used in drug discovery. Collaborations among pharmaceutical institutions can lead to a better performance QSAR prediction, however, intellectual property and related financial interests remain substantially hindering inter-institutional collaborations modeling for discovery.For the first time, we verified feasibility of applying horizontal federated learning (HFL), which recently developed collaborative...
Various computational methods have been developed for quantitative modeling of organic chemical reactions; however, the lack universality as well requirement large amounts experimental data limit their broad applications. Here, we present DeepReac+, an efficient and universal framework prediction reaction outcomes identification optimal conditions based on deep active learning. Under this framework, DeepReac is designed a graph-neural-network-based model, which directly takes 2D molecular...
CRISPR (Clustered Regularly Interspaced Short Palindromic Repeats)-based gene editing has been widely implemented in various cell types and organisms. A major challenge the effective application of system is need to design highly efficient single-guide RNA (sgRNA) with minimal off-target cleavage. Several tools are available for sgRNA design, while limited were compared. In our opinion, benchmarking performance indicating their applicable scenarios important issues. Moreover, whether...
Abstract For genome-wide CRISPR off-target cleavage sites (OTS) prediction, an important issue is data imbalance—the number of true OTS recognized by whole-genome detection techniques much smaller than that all possible nucleotide mismatch loci, making the training machine learning model very challenging. Therefore, computational models proposed for prediction and scoring should be carefully designed properly evaluated in order to avoid bias. In our study, two tools are taken as examples...
Abstract Background The precise characterization of individual tumors and immune microenvironments using transcriptome sequencing has provided a great opportunity for successful personalized cancer treatment. However, the treatment response is often characterized by in vitro assays or bulk transcriptomes that neglect heterogeneity malignant vivo microenvironment, motivating need to use single-cell Methods Here, we present comboSC, computational proof-of-concept study explore feasibility...
Abstract In silico modelling and analysis of small molecules substantially accelerates the process drug development. Representing understanding is fundamental step for various in molecular tasks. Traditionally, these tasks have been investigated individually separately. this study, we presented X-MOL, which applies large-scale pre-training technology on 1.1 billion representation, then, carefully designed fine-tuning was performed to accommodate diverse downstream tasks, including property...
scLearn is a metric learning-based framework with measurement and threshold learned automatically for single-cell assignment.
Systematic evaluation of genome-wide Clustered Regularly Interspaced Short Palindromic Repeats (CRISPR) off-target profiles is a fundamental step for the successful application CRISPR system to clinical therapies. Many experimental techniques and in silico tools have been proposed detecting predicting profiles. These tools, however, not systematically benchmarked. A comprehensive benchmark study an integrated strategy that takes advantage currently available improve predictions are needed....
Abstract Motivation Quantitative structure-activity relationship (QSAR) analysis is commonly used in drug discovery. Collaborations among pharmaceutical institutions can lead to a better performance QSAR prediction, however, intellectual property and related financial interests remain substantially hindering inter-institutional collaborations modeling for Results For the first time, we verified feasibility of applying horizontal federated learning (HFL), which recently developed...
The discovery of CRISPR-Cas systems has paved the way for advanced gene editing tools. However, traditional Cas methods relying on sequence similarity may miss distant homologs and aren't suitable functional recognition. With protein large language models (LLMs) evolving, there is potential system modeling without extensive training data. Here, we introduce CHOOSER (Cas HOmlog Observing SElf-processing scReening), an AI framework alignment-free with self-processing pre-crRNA capability using...
Deciphering taxonomical structures based on high dimensional sequencing data is still challenging in metagenomics study. Moreover, the common workflow processed this field fails to identify microbial communities and their effect a specific disease status. Even relationships interactions between different bacteria community keep unknown. MetaTopics can efficiently extract latent which reflect intrinsic relations or among several major microbes. Furthermore, quantitative measurement, Quetelet...
Abstract The discovery and functional annotation of CRISPR-Cas systems laid the groundwork for development novel CRISPR-based gene editing tools. Traditional similarity- search-based Cas strategies, which rely heavily on local sequence alignment reference homologs, may overlook a significant number remote homologs with limited similarity; it can not be applied directly recognition. With rapid protein large language models (LLMs), foundation are expected to help model without extensive...
To the Editor: CRISPR-based gene editing is widely implemented in various cell types and has great potential for molecular therapy.1Corrigan-Curay J O'Reilly M Kohn DB Cannon PM Bao G Bushman FD et al.Genome technologies: defining a path to clinic.Mol Ther. 2015; 23: 796-806Abstract Full Text PDF PubMed Scopus (83) Google Scholar The CRISPR-Cas9 system creates sequence-specific double-strand DNA breaks that are repaired by dominant error-prone nonhomologous end-joining (NHEJ) pathway, often...
Abstract The identification of T cell neo-epitopes is fundamental and computational challenging in tumor immunotherapy study. As the binding pMHC - receptor (TCR) essential condition for to trigger cytotoxic reactivity, several studies have been proposed predict from perspective pMHC-TCR recognition. However, they often failed with inaccurate prediction a single -TCR pair due highly diverse TCR space. In this study, we novel weakly-supervised learning framework, i . e ., TCRBagger ,...