Aberration-corrected ultrafine analysis of miRNA reads at single-base resolution: a k-mer lattice approach

Abstract

Raw sequencing reads of miRNAs contain machine-made substitution errors, or even insertions and deletions (indels). Although the error rate can be low at 0.1%, precise rectification of these errors is critically important because isoform variation analysis at single-base resolution such as novel isomiR discovery, editing events understanding, differential expression analysis, or tissue-specific isoform identification is very sensitive to base positions and copy counts of the reads. Existing error correction methods do not work for miRNA sequencing data attributed to miRNAs’ length and per-read-coverage properties distinct from DNA or mRNA sequencing reads. We present a novel lattice structure combining kmers, (k – 1)mers and (k + 1)mers to address this problem. The method is particularly effective for the correction of indel errors. Extensive tests on datasets having known ground truth of errors demonstrate that the method is able to remove almost all of the errors, without introducing any new error, to improve the data quality from every-50-reads containing one error to every-1300-reads containing one error. Studies on experimental miRNA sequencing datasets show that the errors are often rectified at the 5′ ends and the seed regions of the reads, and that there are remarkable changes after the correction in miRNA isoform abundance, volume of singleton reads, overall entropy, isomiR families, tissue-specific miRNAs, and rare-miRNA quantities.

Turn ‘noise’ to signal: accurately rectify millions of erroneous short reads through graph learning on edit distances

“…Comparing with state-of-the-art methods including k-mer-methods [13][14][15][16][17][18][19], multiple sequence alignment based methods [20][21][22][23][24], and other methods [25][26][27], our noise2read consistently outperforms under 19 metrics on eight UMI-based wet-lab datasets and five simulated single-end and paired-end datasets constructed in this study. It also has superior performance on eight UMI-based wet-lab datasets and four simulated miRNA datasets established previously in published literature.…”

“…We employed two different processes for generating UMI-based simulation datasets to evaluate the proposed method's performance. The first process was designed to simulate April 5, 2024 27/37 single-end miRNA datasets similarly as worked in the literature [27]. Ideally, each unique read corresponds to a UMI.…”

A curated human cellular microRNAome based on 196 primary cell types

“…Further, category “Assay type” was filtered to only include “miRNA-Seq”, “ncRNA-Seq”, “RNA-Seq,” and the broad unknown category of “OTHER.” This resulted in 6,054 runs that were downloaded using fasterq-dump/fastq-dump of the NCBI SRA Tookit (version 2.9.2) (https://github.com/ncbi/sra-tools) [42]. All runs were evaluated for adapter sequences and any samples with barcodes, unique molecular identifiers (UMIs), or adapter sequences on both ends were not processed (n = 1,870 runs were removed) due to the use of miREC in the processing [43]. Four tissue SRA runs, colon (SRR837824), spleen (SRR6853286), liver (SRR950887) and lymph noted (SRR14130226) were also obtained and processed.…”

“…The miRge3.0 pipeline was run in batches (an average of 11 samples, (-s <samples>)) on two computational clusters (BlueHive, University of Rochester and ARES, Johns Hopkins University) and locally on a PC (with 64-128Gb RAM and 12-40 CPUs) [13]. miRge3.0 default parameters were used along with parameters for miRNA error correction [47] and aligned to miRBase v22.1 [31, 32]. A typical run parameter is as follows: miRge3.0 -s SRAS-file.fastq.gz -a <adapter_sequence> -gff -bam -trf -lib miRge3_Lib -on human -db miRBase -o OutputDir -mEC -ks 20 -ke 20…”

MiR-144 regulates adipogenesis by mediating formation of C/EBPα-FOXO1 protein complex

“…We used the TargetScan prediction tool to predict the miRNAs binding sites within C/EBPα-3'UTR and FoxO1-3'UTR (Fig. 2A), meanwhile, the miRNAs usually have distinct tissue expression patterns (26,27), we further identified the WAT specific one among the miRNAs that targeted C/EBPα and FoxO1. By RT-qPCR detection, we found the miR-144 was the highly expressed miRNA in WAT by comparing with other miRNAs, showed in Fig.…”