GSE102497 Processing Pipeline

Publication

Molecular cell (2018) — PMID 29547715

Dataset

Systematic discovery of RNA binding proteins that control microRNA processing

1. 1

   Sequencing reads from small RNA-seq libraries were first trimmed of adapters using cutadapt

   cutadapt v(Inferred with models/gemini-2.5-flash) GitHub

   $ Bash example

   # Install cutadapt if not already installed
   # conda install -c bioconda cutadapt
   
   # Define input and output file names
   INPUT_READS="small_rna_seq_raw.fastq.gz"
   OUTPUT_READS="small_rna_seq_trimmed.fastq.gz"
   
   # Define the 3' adapter sequence (e.g., Illumina TruSeq Small RNA 3' adapter)
   # This adapter sequence might need to be adjusted based on the specific library preparation kit used.
   ADAPTER_SEQUENCE="TGGAATTCTCGGGTGCCAAGGAACTCCAG"
   
   # Trim adapters, filter for minimum length, and perform quality trimming
   # -a: 3' adapter sequence
   # -m: Minimum read length after trimming (e.g., 18 bp for small RNA)
   # -q: Quality cutoff for trimming (e.g., 20)
   # -o: Output file for trimmed reads
   cutadapt \
     -a "${ADAPTER_SEQUENCE}" \
     -m 18 \
     -q 20 \
     -o "${OUTPUT_READS}" \
     "${INPUT_READS}"

   Copy View on GitHub
2. 2

   Reads were then mapped against a database of repetitive elements derived from RepBase18.05.

   bowtie2 (Inferred with models/gemini-2.5-flash) v2.5.0 (Inferred with models/gemini-2.5-flash) GitHub

   $ Bash example

   # Install bowtie2 (if not already installed)
   # conda install -c bioconda bowtie2
   
   # Download or prepare RepBase18.05.fasta
   # RepBase requires a license for access. This file would contain the sequences of repetitive elements from RepBase 18.05.
   # Example: wget -O RepBase18.05.fasta "https://www.girinst.org/repbase/update/RepBase18.05.fasta.gz" # (Hypothetical URL, actual access requires license)
   # gunzip RepBase18.05.fasta.gz
   
   # Build Bowtie2 index for repetitive elements
   bowtie2-build RepBase18.05.fasta RepBase18.05_index
   
   # Align reads to repetitive elements
   # Assuming single-end reads and 8 threads. Adjust -U for single-end or -1 -2 for paired-end reads.
   bowtie2 -x RepBase18.05_index -U input_reads.fastq.gz -S aligned_to_repeats.sam -p 8
   
   # Optional: Convert SAM to BAM, sort, and index for further processing or filtering
   # samtools view -bS aligned_to_repeats.sam | samtools sort -o aligned_to_repeats.bam
   # samtools index aligned_to_repeats.bam

   Copy View on GitHub
3. 3

   Bowtie version 1.0.0 with parameters -S -q -p 16 -e 100 -l 20 was used to align reads against an index generated from Repbase sequences (Langmead et al., 2009).

   Bowtie v1.0.0 GitHub

   $ Bash example

   bash
   # Install Bowtie (if not already installed)
   # conda install -c bioconda bowtie=1.0.0
   
   # Align reads using Bowtie
   # Assuming 'repbase_index' is the base name for the Bowtie index generated from Repbase sequences
   # Assuming 'reads.fastq' is the input FASTQ file
   # Output will be in SAM format due to -S
   bowtie -S -q -p 16 -e 100 -l 20 repbase_index reads.fastq > alignment.sam
   

   Copy View on GitHub
4. 4

   Reads not mapped to Repbase sequences were aligned to the hg19 genome (UCSC assembly) using bowtie version 1.1.1 with parameters -p 8 -k 1 -m 10 -l 25 --best --chunkmbs 128

   Bowtie v1.1.1 GitHub

   $ Bash example

   # Install Bowtie (if not already installed)
   # conda install -c bioconda bowtie=1.1.1
   
   # Define input and output files
   # INPUT_FASTQ represents 'Reads not mapped to Repbase sequences'
   INPUT_FASTQ="unmapped_reads.fastq"
   OUTPUT_SAM="aligned_to_hg19.sam"
   
   # Define the path to the hg19 Bowtie index
   # Replace /path/to/hg19 with the actual path to your Bowtie index files (e.g., if your index files are hg19.1.ebwt, hg19.2.ebwt, etc., then the prefix is hg19).
   BOWTIE_INDEX_PREFIX="/path/to/hg19"
   
   # Align reads to the hg19 genome using Bowtie
   bowtie -p 8 -k 1 -m 10 -l 25 --best --chunkmbs 128 "${BOWTIE_INDEX_PREFIX}" "${INPUT_FASTQ}" > "${OUTPUT_SAM}"

   Copy View on GitHub
5. 5

   counts of reads for each miRNA were calculated from featureCounts using miRBase annotations

   featureCounts v2.0.6 (Inferred with models/gemini-2.5-flash)

   $ Bash example

   # Install subread (which includes featureCounts)
   # conda install -c bioconda subread=2.0.6
   
   # Download miRBase annotations for Homo sapiens (v22, example)
   # wget ftp://mirbase.org/pub/mirbase/CURRENT/genomes/hsa.gff3 -O miRBase_hsa.gff3
   
   # Example input BAM files (replace with actual paths)
   # Assuming aligned reads are in 'aligned_bams/' directory
   INPUT_BAMS="aligned_bams/sample1.bam aligned_bams/sample2.bam"
   MIRBASE_GFF="miRBase_hsa.gff3"
   OUTPUT_COUNTS="miRNA_counts.txt"
   
   # Calculate counts of reads for each miRNA using featureCounts
   # -a: Annotation file
   # -F GFF3: Specify GFF3 format
   # -t miRNA: Feature type to count (assuming 'miRNA' is the feature type in the GFF3)
   # -g ID: Attribute used to group features into genes (e.g., miRNA ID)
   # -s 0: Unstranded (0), use 1 for stranded, 2 for reverse stranded if known
   # -T 8: Use 8 threads
   # -o: Output file
   featureCounts -a "${MIRBASE_GFF}" -F GFF3 -t miRNA -g ID -s 0 -T 8 -o "${OUTPUT_COUNTS}" ${INPUT_BAMS}

   Copy