sRNAnalyzer—a flexible and customizable small RNA sequencing data analysis pipeline

Abstract

Although many tools have been developed to analyze small RNA sequencing (sRNA-Seq) data, it remains challenging to accurately analyze the small RNA population, mainly due to multiple sequence ID assignment caused by short read length. Additional issues in small RNA analysis include low consistency of microRNA (miRNA) measurement results across different platforms, miRNA mapping associated with miRNA sequence variation (isomiR) and RNA editing, and the origin of those unmapped reads after screening against all endogenous reference sequence databases. To address these issues, we built a comprehensive and customizable sRNA-Seq data analysis pipeline—sRNAnalyzer, which enables: (i) comprehensive miRNA profiling strategies to better handle isomiRs and summarization based on each nucleotide position to detect potential SNPs in miRNAs, (ii) different sequence mapping result assignment approaches to simulate results from microarray/qRT-PCR platforms and a local probabilistic model to assign mapping results to the most-likely IDs, (iii) comprehensive ribosomal RNA filtering for accurate mapping of exogenous RNAs and summarization based on taxonomy annotation. We evaluated our pipeline on both artificial samples (including synthetic miRNA and Escherichia coli cultures) and biological samples (human tissue and plasma). sRNAnalyzer is implemented in Perl and available at: http://srnanalyzer.systemsbiology.net/.

Phospho‐RNA‐seq: a modified small RNA‐seq method that reveals circulating mRNA and lncRNA fragments as potential biomarkers in human plasma

“…Stage 1 of the pipeline involves trimming adapters, removing low‐quality bases, and eliminating reads shorter than 15 nucleotides (see Materials and Methods for additional criteria). Next, we adapted the sRNAnalyzer pipeline (Wu et al , ) to quantify and remove reads aligning to any one of several sequence libraries containing exogenous RNAs (bacterial, fungal, and viral), various endogenous non‐coding RNA sequences, and other possible contaminants (transposons, repetitive elements, and UniVec contaminants; Stage 2). Reads with no valid alignments to these sequence libraries in Stage 2 are then aligned to the human genome.…”

“…TruSeq adapters and stop oligo sequences were trimmed with cutadapt (v 1.91) using processing steps adapted from the sRNAnalyzer workflow (Martin, ; Wu et al , ). The sRNAnalyzer framework was also adapted to align adapter‐trimmed reads 15 nt and longer to several sequence databases containing known small RNA families and contaminant sequences (Wu et al , ). A table with descriptions of the included sequence databases is provided in Appendix Table S2.…”

MicroRNA‐transcriptome networks in whole blood and monocytes of women undergoing preterm labour

“…Our analysis stands out among related studies because it is the first to generate high throughput RNA sequencing data of both miRNA and mRNA from the same patient source. We have carefully selected a reliable methodology to identify potential miRNA targets based on experimental data and have used novel alignment tools to obtain comprehensive coverage of miRNA expression . This robust, multi‐dimensional data have allowed us to confidently build mRNA‐miRNA regulation networks related to sPTL in a way that has not previously been done, which provides additional insight into mechanisms of transcriptional regulation of human labour.…”

“…We have carefully selected a reliable methodology to identify potential miRNA targets based on experimental data 23 and have used novel alignment tools to obtain comprehensive coverage of miRNA expression. 20 This robust, multi-dimensional data have allowed us to confidently build mRNA-miRNA regulation networks related to sPTL in a way that has not previously been done, which provides additional insight into mechanisms of transcriptional regulation of human labour. This type of integrative analysis has been suggested to overcome many challenges in performing pregnancy research.…”

Extracellular microRNA 3’ end modification across diverse body fluids

“…1A). While other methods to identify miRNA sequence variations exist [28][29][30][31][32][33][34][35][36] , our method aims to improve the mapping strategies and reduce false positive modifications (Methods). Briefly, in the mapping steps, we sequentially trim the ends of unmapped reads (separately for 3' and 5' ends) and check for one or more consecutive NTAs.…”