<b>VeloRM: Inferring single-cell RNA modification velocity by disentangling pre- and post-splicing epitranscriptomes</b>

RNA modification plays a critical role in regulating the function and fate of RNA molecules, yet its dynamics across the RNA life cycle remain under-explored. We present here VeloRM, a computational framework that disentangles the pre-splicing and post-splicing epitranscriptomes, enabling the inference of RNA modification kinetics and the prediction of future epitranscriptome states at single-nucleotide and single-cell resolution. Applying VeloRM to three single-cell epitranscriptome datasets of m⁶A and A-to-I, we uncover distinct modification patterns on the pre-splicing and post-splicing RNAs. Notably, N⁶-methyladenosine (m⁶A) is more heavily decorated on pre-splicing RNAs than post-splicing RNAs near the splicing junctions. Importantly, VeloRM rigorously quantifies the velocity of RNA modification, provides direct insight into the future epitranscriptome landscape of individual cells, aligning well with known data on single-cell trajectories within the cell cycle and during cell differentiation. Together, our study broadens the current understanding of the dynamic epitranscriptome by establishing a rigorous framework for tracing its kinetics across the RNA life cycle in a dynamic process spanning the pre-splicing and post-splicing RNA states.<br><b>Code</b>The repository includes R Markdown scripts for analyzing both <b>scDART-seq data</b> and <b>RNA-Seq A-to-I editing data</b>.<b>scDART-seq Data Analysis</b>The analysis is divided into three main stages: <b>(1) raw data preprocessing</b>, <b>(2) data processing</b>, and <b>(3) downstream analysis</b>.<b>1. Raw Data (origin)</b><b>SNP_all</b>: SNP information for candidate m6A sites.<b>frequency_pre_YTH</b>: Methylated and unmethylated read counts in pre-mRNA of test (YTH-WT) cells.<b>frequency_pre_YTHmut</b>: Methylated and unmethylated read counts in pre-mRNA of control (YTH-mutant) cells.<b>frequency_spliced_YTH</b>: Methylated and unmethylated read counts in mature mRNA of test cells.<b>frequency_spliced_YTHmut</b>: Methylated and unmethylated read counts in mature mRNA of control cells.<b>frequency_isoform_YTH</b>: Methylated and unmethylated read counts in ambiguous (isoform-level) mRNA of test cells.<b>frequency_isoform_YTHmut</b>: Methylated and unmethylated read counts in ambiguous mRNA of control cells.<b>label_SigRM</b>: Cell cycle stage labels from <i>Statistical modeling of single-cell epitranscriptomics enabled trajectory and regulatory inference of RNA methylation</i> (Cell Genomics).<b>expression_TPM</b>: Gene expression data (TPM) from the same study.<b>gene_informations</b>: Gene annotation data from the same study.<b>2. Preprocessing</b><b>SNP_37199</b>: Subset of SNPs used for downstream m6A analysis.<b>test_list</b>: Processed read counts for test cells, including:<code>spliced_meth_test</code>, <code>spliced_unmeth_test</code><code>pre_meth_test</code>, <code>pre_unmeth_test</code><code>isoform_meth_test</code>, <code>isoform_unmeth_test</code><b>control_list</b>: Processed read counts for control cells, similarly structured.<b>res_preprocess</b>: Preprocessed input formatted for VeloRM.<b>res_DESeq2</b>: Differential expression analysis results using DESeq2.<b>SigRMtest_res_unspliced</b>: SigRM test results on unspliced reads (test vs. control).<b>SigRMtest_res_spliced</b>: SigRM test results on spliced reads (test vs. control).<b>3. Analysis</b><b>splicing_junction_analysis</b>: Data frame with distances between SNPs and the nearest splicing junctions.<b>res_m6A_velocity</b>: m6A velocity results computed by VeloRM.<b>res_aboundance_velocity</b>: Site-level dboundance velocity estimated by VeloRM.<b>res_transcriptional_impact_analysis_spearman</b>: Transcriptional impact results (Spearman correlation-based) from VeloRM.<b>cell_cycle_sites</b>: Cell cycle-related methylation site analysis performed with VeloRM.<b>RNA-Seq A-to-I Data Analysis</b><b>1. Raw Data (origin)</b><b>SNP_all</b>: SNP information of candidate A-to-I editing sites.<b>pre_A-to-I</b>: Read counts (methylated and unmethylated) in pre-mRNA of test cells.<b>spliced_A_to_I</b>: Read counts in mature mRNA of test cells.<b>isoform_A_to_I</b>: Read counts in ambiguous mRNA of test cells.<b>2. Preprocessing</b><b>SNP_1159</b>: Filtered SNPs for downstream A-to-I analysis.<b>test_list</b>: Read counts for test cells, including:<code>spliced_meth_test</code>, <code>spliced_unmeth_test</code><code>pre_meth_test</code>, <code>pre_unmeth_test</code><code>isoform_meth_test</code>, <code>isoform_unmeth_test</code><b>control_list</b>: <code>NULL</code> (no control group in this dataset).<b>res_preprocess</b>: Preprocessed data formatted for VeloRM.<b>res_DESeq2</b>: Differential analysis results via DESeq2.<b>3. Analysis</b><b>splicing_junction_analysis</b>: Distances between candidate sites and nearest splice junctions.<b>cell_cycle_sites</b>: Cell cycle-related A-to-I site analysis via VeloRM.<b>RNA-Seq A-to-I Data Analysis 2</b><b>1. Raw Data (origin)</b><b>SNP_all</b>: SNP information of candidate A-to-I editing sites.<b>pre_A-to-I</b>: Read counts (methylated and unmethylated) in pre-mRNA of test cells.<b>spliced_A_to_I</b>: Read counts in mature mRNA of test cells.<b>isoform_A_to_I</b>: Read counts in ambiguous mRNA of test cells.<b>2. Preprocessing</b><b>SNP_1931</b>: Filtered SNPs for downstream A-to-I analysis.<b>test_list</b>: Read counts for test cells, including:<code>spliced_meth_test</code>, <code>spliced_unmeth_test</code><code>pre_meth_test</code>, <code>pre_unmeth_test</code><code>isoform_meth_test</code>, <code>isoform_unmeth_test</code><b>control_list</b>: <code>NULL</code> (no control group in this dataset).<b>res_preprocess</b>: Preprocessed data formatted for VeloRM.<b>res_DESeq2</b>: Differential analysis results via DESeq2.<b>3. Analysis</b><b>splicing_junction_analysis</b>: Distances between candidate sites and nearest splice junctions.<b>cell_cycle_sites</b>: Cell cycle-related A-to-I site analysis via VeloRM.<b>DART-seq Data Analysis</b><b>1. Raw Data (origin)</b><b>SNP_all</b>: SNP information of candidate A-to-I editing sites.<b>pre_m6A</b>: Read counts (methylated and unmethylated) in pre-mRNA of test cells.<b>spliced_</b><b>m6A</b>: Read counts in mature mRNA of test cells.<b>isoform_</b><b>m6A</b>: Read counts in ambiguous mRNA of test cells.<b>2. Preprocessing</b><b>SNP</b>: Filtered SNPs for downstream A-to-I analysis.<b>test_list</b>: Read counts for test cells, including:<code>spliced_meth_test</code>, <code>spliced_unmeth_test</code><code>pre_meth_test</code>, <code>pre_unmeth_test</code><code>isoform_meth_test</code>, <code>isoform_unmeth_test</code><b>control_list</b>: <code>NULL</code> (no control group in this dataset).<b>res_preprocess</b>: Preprocessed data formatted for VeloRM.<b>res_DESeq2</b>: Differential analysis results via DESeq2.<b>res_DESeq2_pre</b>: Differential analysis results via DESeq2.<b>res_DESeq2_spliced</b>: Differential analysis results via DESeq2.<b>3. Analysis</b><b>splicing_junction_analysis</b>: Distances between candidate sites and nearest splice junctions.<b>cell_cycle_sites</b>: Cell cycle-related A-to-I site analysis via VeloRM.