2.7 Sequencing data analysis
Raw sequencing data were processed to remove adaptors and low-quality bases, and high-quality reads were aligned with the rice reference genome. The sequenced read coverage at 3′-UTR, coding sequence (CDS), and 5′-UTR of the target transcripts in each library were calculated and normalized. TPM (Transcripts Per Kilobase Million) was used as a measure of transcript expression levels. In the TPM calculation process, the reads count value for each transcript was divided by its transcript length (in kilobases) to obtain the reads per kilobase coverage (RPK) of the transcript. The RPK values for all transcripts are summed and divided by 1,000,000 to obtain a million-scaling factor. Subsequently, the RPK values are divided by the million-scaling factor to obtain the TPM values. The sum of all TPM values in each sample remains constant, allowing for easier comparison of the proportion of reads mapped to transcripts in each sample. The expression quantification of transcripts was performed using salmon (version: 1.4.0) (Patro et al., 2017), and then subjected to differential expression analysis. DESeq2 (version: 1.26.0) was used as the software for differential expression analysis, with a threshold of padj (p -value) < 0.05 and |log2FoldChange| > 1. MINES pipeline (https://github.com/YeoLab/ MINES) (Lorenz et al., 2020) was used to identify m6A modification sites in RNA sequences. Meme (version: 5.3.3) (Bailey et al., 2009) was used to extend two bases up- and down-stream of the m6A methylation sites, resulting in a motif comprising five bases. Methylkit (Akalin et al., 2012) was used for differential methylation loci analysis. Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway enrichment analyses of the differentially expressed genes were performed using the GOseq R package and KOBAS 3.0 software, respectively (Ashburner et al., 2000). The enrichment analysis was performed based on the hypergeometric test (Kanehisa et al., 2008).