Biology subjects

Yuan, Y.

Publications and source records attributed to Yuan, Y..

At least 163 records · Page 9Linked to original sources

bioRxiv2024

Deciphering the diversity and sequence of extracellular matrix and cellular spatial patterns in lung adenocarcinoma using topological data analysis

Extracellular matrix (ECM) organization influences cancer development and progression. It modulates the invasion of cancer cells and can hinder the access of immune cells to cancer cells. Effective quantification of ECM architecture and its relationship to the position of different cell types is, therefore, important when investigating the role of ECM in cancer development. Using topological data analysis (TDA), particularly persistent homology and Dowker persistent homology, we develop a novel analysis pipeline for quantifying ECM architecture, spatial patterns of cell positions, and the spatial relationships between distinct constituents of the tumour microenvironment. We apply the pipeline to 44 surgical specimens of lung adenocarcinoma from the lung TRACERx study stained with picrosirius red and haematoxylin. We show that persistent homology effectively encodes the architectural features of the tumour microenvironment. Inference using pseudo-time analysis and spatial mapping to centimetre scale tissues suggests a gradual and progressive route of change in ECM architecture, with two different end states. Dowker persistent homology enables the analysis of spatial relationship between any pair of constituents of the tumour microenvironment, such as ECM, cancer cells, and leukocytes. We use Dowker persistent homology to quantify the spatial segregation of cancer and immune cells over different length scales. A combined analysis of both topological and non-topological features of the tumour microenvironment indicates that progressive changes in the ECM are linked to increased immune exclusion and reduced oxidative metabolism.

cancer biology↗

bioRxiv2024

Insights into Human Epileptogenesis with Proteomic Profiling

Epilepsy affects millions globally, and drug-resistant epilepsy remains a challenge. Molecular mechanisms underlying epilepsy remain elusive. Protein profiling through proteomics offers insight into biomarkers and therapeutic targets. Human brain tissue from epilepsy surgeries was analyzed using data-independent acquisition (DIA) proteomics. Samples were categorized into Core (epileptogenic focus), Border (marginal excision tissue), and Nonepileptic control groups. Differential expression proteins (DEPs) were identified and shared proteins were analyzed. 163 DEPs were identified which may has potential roles in the initiation of epileptic electrical firing, 412 DEPs which indicating the difference between epilepsy and Nonepilepsy patients and 10 DEPs consistently altered in Core which indicating potential roles in epileptogenesis. Notably, P35754/GLRX, O75335/PPFIA4, and Q96KP4/CNDP2 were consistently expressed differently in all group pairs. From validation experiments, the expression of Kv3.2 significant reduced in the Core group compare to border group by immunohistochemistry and knockdown of Kv3.2 increased seizure susceptibility and altered neuronal excitability through our cellular and animal experimentation.

neuroscience↗

bioRxiv2024

Development of deaminase-free T-to-S base editor and C-to-G base editor by engineered human uracil DNA glycosylase

DNA base editors could enable direct editing of adenine (A), cytosine (C), or guanine (G), but there is no base editor for direct thymine (T) editing currently. Here, by fusing Cas9 nickase (nCas9) with engineered human uracil DNA glycosylase (UNG) variants, we developed a deaminase-free glycosylase-based thymine base editor (gTBE) with the ability of direct T editing. By several rounds of UNG mutagenesis via rational screening, we demonstrated that gTBE with engineered UNG variants could achieve T editing efficiency by up to 81.5%. Furthermore, the gTBE exhibited high T-to-S (i.e., T-to-C or T-to-G) conversion ratio with up to 0.97 in cultured human cells. Using similar strategy, we developed a deaminase-free cytosine base editor (gCBE) facilitating specifically direct C editing by engineered UNG with mutations different from gTBE. Thus, we provide two novel base editors, gTBE and gCBE, with corresponding engineered UNG variants, broadening the targeting scope of base editors.

molecular biology↗

bioRxiv2023

An Efficient and Safe Trans-complementation System of Mpox Virus Reproduces Authentic Viral Architecture and Infection

Since the Mpox virus (MPXV) outbreak in 2022, there have been 97,745 cases and 203 fatalities. This outbreak features increased transmission efficiency and a higher infection rate in the MSM population, with the underlying causes remaining unknown. The requirement of BSL-3 laboratories poses a barrier to MPXV research and antiviral development. Here, we report an efficient and safe trans-complementary system that produces a single-round infectious MPXV, preserving the viruss authentic architecture and enabling it to complete its life cycle in complementing cells. This deficient MPXV genome, lacking two essential genes crucial for late gene transcription and viral assembly, is restricted to a single-round infection in non-complementing cell lines. Notably, SCID mice inoculated with this deficient MPXV exhibited no detectable disease or viral load. This deficient MPXV platform has demonstrated its capacity to study innate immunity and cell death during infection in complementing cells. It can also be used for antibody neutralization assays and anti-MPXV drug evaluation. This trans-complementation platform, safe for use in low-biosafety laboratories, offers a valuable resource for MPXV research and countermeasure development.

microbiology↗

bioRxiv2023

Queuosine Salvage in Bartonella henselae Houston 1: A Unique Evolutionary Path

Queuosine (Q) stands out as the sole tRNA modification that can be synthesized via salvage pathways. Comparative genomic analyses identified specific bacteria that showed a discrepancy between the projected Q salvage route and the predicted substrate specificities of the two identified salvage proteins: 1) the distinctive enzyme tRNA guanine-34 transglycosylase (bacterial TGT, or bTGT), responsible for inserting precursor bases into target tRNAs; and 2) Queuosine Precursor Transporter (QPTR), a transporter protein that imports Q precursors. Organisms like the facultative intracellular pathogen Bartonella henselae, which possess only bTGT and QPTR but lack predicted enzymes for converting preQ1 to Q, would be expected to salvage the queuine (q) base, mirroring the scenario for the obligate intracellular pathogen Chlamydia trachomatis. However, sequence analyses indicate that the substrate-specificity residues of their bTGTs resemble those of enzymes inserting preQ1 rather than q. Intriguingly, mass spectrometry analyses of tRNA modification profiles in B. henselae reveal trace amounts of preQ1, previously not observed in a natural context. Complementation analysis demonstrates that B. henselae bTGT and QPTR not only utilize preQ1, akin to their Escherichia coli counterparts, but can also process q when provided at elevated concentrations. The experimental and phylogenomic analyses suggest that the Q pathway in B. henselae could represent an evolutionary transition among intracellular pathogens--from ancestors that synthesized Q de novo to a state prioritizing the salvage of q. Another possibility that will require further investigations is that the insertion of preQ1 has fitness advantages when B. henselae is growing outside a mammalian host. Author summaryTransfer RNAs (tRNAs) are adaptors that deliver amino acids to ribosomes during translation of messenger RNAs (mRNAs) into proteins. tRNA molecules contain specially-modified nucleotides that affect many aspects of translation, including regulation of translational efficiency, as modified nucleotides primarily occur near the portion of tRNA (anticodon) that directly interacts with the coding sequence (codon) of the mRNA while it is associated with a ribosome. Queuosine (Q) is a modified tRNA nucleotide located in the anticodon that can be synthesized or uniquely imported from the environment as Q or a precursor using a salvage mechanism. Free-living bacteria, e.g., E. coli, can synthesize Q or salvage precursors from the environment, but many obligate intracellular pathogens, e.g., Chlamydia trachomatis, cannot synthesize Q and must import a precursor from eukaryotic hosts. In this study, we determined that Bartonella henselae, a facultative intracellular bacterial pathogen of vascular cells, falls somewhere in the middle, as it is unable to synthesize Q but can salvage Q or certain precursors. The unusual nature of Bartonellas system suggests different evolutionary scenarios. It could be a snapshot of the transition from Q synthesis to strict Q salvage or represent a unique adaptation to a complex multi-host lifestyle.

microbiology↗

bioRxiv2023

Integration of multi-source gene interaction networks and omics data with graph attention networks to identify novel disease genes

The pathogenesis of diseases is closely associated with genes, and the discovery of disease genes holds significant importance for understanding disease mechanisms and designing targeted therapeutics. However, biological validation of all genes for diseases is expensive and challenging. In this study, we propose DGP-AMIO, a computational method based on graph attention networks, to rank all unknown genes and identify potential novel disease genes by integrating multi-omics and gene interaction networks from multiple data sources. DGP-AMIO outperforms other methods significantly on 20 disease datasets, with an average AUROC and AUPR exceeding 0.9. The superior performance of DGP-AMIO is attributed to the integration of multiomics and gene interaction networks from multiple databases, as well as triGAT, a proposed GAT-based method that enables precise identification of disease genes in directed gene networks. Enrichment analysis conducted on the top 100 genes predicted by DGP-AMIO and literature research revealed that a majority of enriched GO terms, KEGG pathways and top genes were associated with diseases supported by relevant studies. We believe that our method can serve as an effective tool for identifying disease genes and guiding subsequent experimental validation efforts.

bioinformatics↗

bioRxiv2023

AttSiOff: A self-attention-based approach on siRNA de-sign with inhibition and off-target effect prediction

MotivationSmall interfering RNA (siRNA) is often used for function study and expression regulation of specific genes, as well as the development of small molecule drugs. Selecting siRNAs with high inhibition and low off-target effects from massive candidates is always a great challenge. Increasing experimentally validated samples prompt the development of machine-learning-based algorithms, including Support Vector Machine (SVM), Convolutional Neural Network (CNN), and Graph Neural Network (GNN). However, these methods still suffer from limited accuracy and poor generalization to design both potent and specific siRNAs. ResultsIn this study, we propose a novel approach for siRNA inhibition and off-target effect prediction, named AttSiOff. It combines self-attention-based siRNA inhibition predictor with an mRNA searching package and an off-target filter. The predictor gives the inhibition score via analyzing the embedding of siRNA and local mRNA sequences, generated from pre-trained RNA-FM model, as well as other meaningful prior-knowledge-based features. Self-attention mechanism can detect potentially decisive features, which may determine the inhibition of siRNA. It captures global and local dependencies more efficiently than normal convolutions. The 10-fold cross-validation results indicate that our model achieves a significant improvement of correlation between prediction and label, compared with all existing methods. And it reaches better performance of generalization and robustness on cross-dataset validation. In addition, the mRNA searching package could find all mature mRNAs for given gene name from GENOMES database, and the off-target filter can calculate the amount of unwanted off-target binding sites, which affects the specificity of siRNA. Experiments on mature siRNA drugs show that our entire framework, AttSioff, have excellent convenience and operability in practical applications. Contactyuanye_auto@sjtu.edu.cn or chengjin520@sjtu.edu.cn.

bioinformatics↗

bioRxiv2023

Identification of functional genes underlining growth in large yellow croaker

Large yellow croaker (Larimichthys crocea) is the largest cultured marine fish in China, although its economic benefits are significant, its growth cycle is long and it takes 2 years to market. Identification of genomic variations and genes affecting the growth of L. crocea is of great significance for breeding fast-growing strains by molecular means. We utilized whole genome resequencing to identify a highly effective GWAS locus on chromosome 6 of the genome, and identified some candidate genes that affect growth. This study provides a theoretical basis for analyzing the genetic mechanism of growth traits of L. crocea and cultivating fast-growing varieties.

genetics↗

bioRxiv2023

Functional Screening in human HSPCs identifies optimized protein-based enhancers of Homology Directed Repair

Homology Directed Repair (HDR) enables precise genome editing and holds great promise in the gene therapy field. However, the implementation of HDR-based therapies is hindered by limited efficiency in comparison to methods that exploit alternative DNA repair routes, such as Non-Homologous End Joining (NHEJ). In this study, we demonstrate the development of a functional, pooled screening platform utilizing an HDR-based readout to identify protein-based reagents that improve HDR outcomes in human hematopoietic stem and progenitor cells (HSPCs), a clinically relevant cell type for gene therapy. We leveraged this screening platform to explore sequence diversity at the binding interface of the NHEJ inhibitor i53 and its target, 53BP1, and we identified optimized i53 variants that enable new intermolecular bonds and robustly increase HDR. These variants specifically reduce insertion-deletion outcomes and also synergize with a DNAPK inhibitor to increase HDR rates. When applied at manufacturing scale, the incorporation of improved variants results in a significant increase in cells with at least one repaired allele and improved HDR in long-term HSPCs subpopulations, while not increasing off-target editing or gross chromosomal rearrangements. We anticipate the pooled screening platform will enable discovery of future gene editing reagents that improve HDR outcomes, such as the i53 variants reported here.

molecular biology↗

bioRxiv2023

Cross-modal Generation of Hit-like Molecules via Foundation Model Encoding of Gene Expression Signatures

Designing de novo molecules with specific biological activity is an essential task since it holds the potential to bypass the exploration of target genes, which is an initial step in the modern drug discovery paradigm. However, traditional methods mainly screen molecules by comparing the desired molecular effects within the documented experimental results. The data set limits this process, and it is hard to conduct direct cross-modal comparisons. Therefore, we propose a solution based on cross-modal generation called GexMolGen (Gene Expression-based Molecule Generator), which generates hit-like molecules using gene expression signatures alone. These signatures are calculated by inputting control and desired gene expression states. Our model GexMolGen adopts a "first-align-then-generate" strategy, aligning the gene expression signatures and molecules within a mapping space, ensuring a smooth cross-modal transition. The transformed molecular embeddings are then decoded into molecular graphs. In addition, we employ an advanced single-cell large language model for input flexibility and pre-train a scaffold-based molecular model to ensure that all generated molecules are 100% valid. Empirical results show that our model can produce molecules highly similar to known references, whether feeding in- or out-of-domain transcriptome data. Furthermore, it can also serve as a reliable tool for cross-modal screening.

bioinformatics↗

bioRxiv2023

DeepSipred: A deep-learning-based approach on siRNA inhibition prediction

MotivationThe use of exogenous small interfering RNAs (siRNAs) for gene silencing has become a widespread molecular tool for gene function study and new drug identification. Although the pathway of RNAi to mediate gene expression has been widely investigated, the selection of hyperfunctional siRNA with high inhibition remains challenging. ResultsIn this study, we build a deep-learning-based approach on siRNA inhibition prediction, named DeepSipred. It combines features from sequence context, thermodynamic property, and other expert knowledge together to predict the inhibition more accurately than existing methods. The sequence features from siRNA and local target mRNA are generated via one-hot encoding and pretrained RNA-FM encoding. The convolution layers with multiple kernels in DeepSipred can detect various decisive motifs, which will determine the actual inhibition of siRNA. The thermodynamic features are calculated from Gibbs Free Energy. In addition, the expert knowledge includes those design criteria from previous studies. Benchmarked on large available public datasets, the 10-fold cross-validation results indicate that our predictor achieving the state-of-the-art performance. Contactyuanye_auto@sjtu.edu.cn or chengjin520@sjtu.edu.cn.

bioinformatics↗

bioRxiv2023

A novel Bayesian model for assessing intratumor heterogeneity of tumor infiltrating leukocytes with multi-region gene expression sequencing

Intratumor heterogeneity (ITH) of tumor-infiltrated leukocytes (TILs) is an important phenomenon of cancer biology with potentially profound clinical impacts. Multiregion gene expression sequencing data provide a promising opportunity that allows for explorations of TILs and their intratumor heterogeneity for each subject. Although several existing methods are available to infer the proportions of TILs, considerable methodological gaps exist for evaluating intratumor heterogeneity of TILs with multi-region gene expression data. Here, we develop ICeITH, immune cell estimation reveals intratumor heterogeneity, a Bayesian hierarchical model that borrows cell type profiles as prior knowledge to decompose mixed bulk data while accounting for the within-subject correlations among tumor samples. ICeITH quantifies intratumor heterogeneity by the variability of targeted cellular compositions. Through extensive simulation studies, we demonstrate that ICeITH is more accurate in measuring relative cellular abundance and evaluating intratumor heterogeneity compared with existing methods. We also assess the ability of ICeITH to stratify patients by their intratumor heterogeneity score and associate the estimations with the survival outcomes. Finally, we apply ICeITH to two multi-region gene expression datasets from lung cancer studies to classify patients into different risk groups according to the ITH estimations of targeted TILs that shape either pro- or anti-tumor processes. In conclusion, ICeITH is a useful tool to evaluate intratumor heterogeneity of TILs from multi-region gene expression data.

bioinformatics↗

bioRxiv2023

Automated, self-resistance gene-guided, and high-throughput genome mining of bioactive natural products from Streptomyces

Natural products (NPs) produced by bacteria, fungi and plants are a major source of drug leads. Streptomyces species are particularly important in this regard as they produce numerous natural products with prominent bioactivities. Here we report a fully automated, scalable and high-throughput platform for discovery of bioactive natural products in Streptomyces (FAST-NPS). This platform comprises computational prediction and prioritization of target biosynthetic gene clusters (BGCs) guided by self-resistance genes, highly efficient and automated direct cloning and heterologous expression of BGCs, followed by high-throughput fermentation and product extraction from Streptomyces strains. As a proof of concept, we applied this platform to clone 105 BGCs ranging from 10 to 100 kb that contain potential self-resistance genes from 11 Streptomyces strains with a success rate of 95%. Heterologous expression of all successfully cloned BGCs in Streptomyces lividans TK24 led to the discovery of 23 natural products from 12 BGCs. We selected 5 of these 12 BGCs for further characterization and found each of them could produce at least one natural product with antibacterial and/or anti-tumor activity, which resulted in a total of 8 bioactive natural products. Overall, this work would greatly accelerate the discovery of bioactive natural products for biomedical and biotechnological applications. Graphic Abstracts O_FIG O_LINKSMALLFIG WIDTH=200 HEIGHT=184 SRC="FIGDIR/small/564101v1_ufig1.gif" ALT="Figure 1"> View larger version (41K): org.highwire.dtl.DTLVardef@11068c9org.highwire.dtl.DTLVardef@4f68c5org.highwire.dtl.DTLVardef@16767edorg.highwire.dtl.DTLVardef@1d82eab_HPS_FORMAT_FIGEXP M_FIG C_FIG

synthetic biology↗

bioRxiv2023

Modulation of Allostery with Multiple Mechanisms by Hotspot Mutations in TetR

Modulating allosteric coupling offers unique opportunities for biomedical applications. Such efforts can benefit from efficient prediction and evaluation of allostery hotspot residues that dictate the degree of co-operativity between distant sites. We demonstrate that effects of allostery hotspot mutations can be evaluated qualitatively and semi-quantitatively by molecular dynamics simulations in a bacterial tetracycline repressor (TetR). The simulations recapitulate the effects of these mutations on abolishing the induction function of TetR and provide a rationale for the different degrees of rescuability observed to restore allosteric coupling of the hotspot mutations. We demonstrate that the same non-inducible phenotype could be the result of perturbations in distinct structural and energetic properties of TetR. Our work underscore the value of explicitly computing the functional free energy landscapes to effectively evaluate and rank hotspot mutations despite the prevalence of compensatory interactions, and therefore provide quantitative guidance to allostery modulation for therapeutic and engineering applications. O_FIG O_LINKSMALLFIG WIDTH=200 HEIGHT=111 SRC="FIGDIR/small/555381v3_ufig1.gif" ALT="Figure 1"> View larger version (19K): org.highwire.dtl.DTLVardef@16d5ddeorg.highwire.dtl.DTLVardef@12da7e3org.highwire.dtl.DTLVardef@693d33org.highwire.dtl.DTLVardef@1311afa_HPS_FORMAT_FIGEXP M_FIG O_FLOATNOTOC GraphicC_FLOATNO C_FIG

biophysics↗

bioRxiv2023

Mapping Human Tissues with Highly Multiplexed RNA in situ Hybridization

In situ transcriptomic techniques promise a holistic view of tissue organization and cell-cell interactions. Recently there has been a surge of multiplexed RNA in situ techniques but their application to human tissues and clinical biopsies has been limited due to their large size, general lower tissue quality and high background autofluorescence. Here we report DART-FISH, a versatile padlock probe-based technology capable of profiling hundreds to thousands of genes in centimeter-sized human tissue sections at cellular resolution. We introduced an omni-cell type cytoplasmic stain, dubbed RiboSoma that substantially improves the segmentation of cell bodies. We developed a computational decoding-by-deconvolution workflow to extract gene spots even in the presence of optical crowding. Our enzyme-free isothermal decoding procedure allowed us to image 121 genes in a large section from the human neocortex in less than 10 hours, where we successfully recapitulated the cytoarchitecture of 20 neuronal and non-neuronal subclasses. Additionally, we demonstrated the detection of transcripts as short as 461 nucleotides, including neuropeptides and discovered new cortical layer markers. We further performed in situ mapping of 300 genes on a diseased human kidney, profiled >20 healthy and pathological cell states, and identified diseased niches enriched in transcriptionally altered epithelial cells and myofibroblasts.

genomics↗

bioRxiv2023

Clock-dependent chromatin accessibility rhythms regulate circadian transcription

Chromatin organization plays a crucial role in gene regulation by controlling the accessibility of DNA to transcription machinery. While significant progress has been made in understanding the regulatory role of clock proteins in circadian rhythms, how chromatin organization affects circadian rhythms remains poorly understood. Here, we employed ATAC-seq (Assay for Transposase-Accessible Chromatin with Sequencing) on FAC-sorted Drosophila clock neurons to assess genome-wide chromatin accessibility over the circadian cycle. We observed significant circadian oscillations in chromatin accessibility at promoter and enhancer regions of hundreds of genes, with enhanced accessibility either at dusk or dawn, which correlated with their peak transcriptional activity. Notably, genes with enhanced accessibility at dusk were enriched with E-box motifs, while those more accessible at dawn were enriched with VRI/PDP1-box motifs, indicating that they are regulated by the core circadian feedback loops, PER/CLK and VRI/PDP1, respectively. Further, we observed a complete loss of chromatin accessibility rhythms in per01 null mutants, with chromatin consistently accessible throughout the circadian cycle, underscoring the critical role of Period protein in driving chromatin compaction during the repression phase. Together, this study demonstrates the significant role of chromatin organization in circadian regulation, revealing how the interplay between clock proteins and chromatin structure orchestrates the precise timing of biological processes throughout the day. This work further implies that variations in chromatin accessibility might play a central role in the generation of diverse circadian gene expression patterns in clock neurons. Significance StatementChromatin organization plays a critical role in gene regulation in development and in disease. In this study, we discovered robust circadian oscillations in the chromatin accessibility of regulatory elements of clock-regulated genes in Drosophila clock neurons, with enhanced accessibility either at dusk or dawn, which correlated with their peak transcriptional activity. We found enrichment of E-box motifs in genes that exhibited enhanced accessibility at dusk, and enrichment of VRI/PDP1-box motifs in genes that exhibited higher accessibility at dawn. Moreover, the complete loss of chromatin accessibility rhythms in per01 mutants highlights the essential role of the Period protein in driving chromatin compaction during the repression phase. This study highlights the significance of chromatin organization in the generation of [~]24-hour circadian rhythms.

neuroscience↗

bioRxiv2023

Histone variant H2A.Z is required for plant salt response by regulating gene transcription

As a well-conserved histone variant, H2A.Z epigenetically regulates plant growth and development as well as the interaction with environmental factors. However, the role of H2A.Z in response to salt stress remains unclear, and whether nucleosomal H2A.Z occupancy work on the gene responsiveness upon salinity is obscure. Here, we elucidate the involvement of H2A.Z in salt response by analyzing H2A.Z disorder plants with impaired or overloaded H2A.Z deposition. The salt tolerance is dramatically accompanied by H2A.Z deficiency and reacquired in H2A.Z OE lines. H2A.Z disorder changes the expression profiles of large-scale of salt responsive genes, announcing that H2A.Z is required for plant salt response. Genome-wide H2A.Z mapping shows that H2A.Z level is induced by salt condition across promoter, TSS and TES (-1 kb to +1kb), the peaks preferentially enrich at promoter regions near TSS. We further show that H2A.Z deposition within TSS provides a direct role on transcriptional control, which has both repressive and activating effects, while it is found generally H2A.Z enrichment negatively correlate with gene expression level response to salt stress. This study shed light on the H2A.Z function in salt tolerance, highlighting the complex regulatory mechanisms of H2A.Z on transcriptional activity for yielding appropriate responses to particularly environmental stress.

plant biology↗

bioRxiv2023

TFvelo: gene regulation inspired RNA velocity estimation

RNA velocity is closely related with cell fate and is an important indicator for the prediction of cell states with elegant physical explanation derived from single-cell RNA-seq data. Most existing RNA velocity models aim to extract dynamics from the phase delay between unspliced and spliced mRNA for each individual gene. However, unspliced/spliced mRNA abundance may not provide sufficient signal for dynamic modeling, leading to poor fit in phase portraits. Motivated by the idea that RNA velocity could be driven by the transcriptional regulation, we propose TFvelo, which expands RNA velocity concept to various single-cell datasets without relying on splicing information, by introducing gene regulatory information. Our experiments on synthetic data and multiple scRNA-Seq datasets show that TFvelo can accurately fit genes dynamics on phase portraits, and effectively infer cell pseudo-time and trajectory from RNA abundance data. TFvelo opens a novel, robust and accurate avenue for modeling RNA velocity for single cell data.

bioinformatics↗