4.3 Raw data cleaning¶

Purpose¶

Remove adapters, discard very short reads, and prepare clean FASTQ files for read classification and alignment.

The workflow is divided into two steps:

Step 1: Clean RNA-seq raw reads
Step 2: Clean Ribo-seq raw reads

Each step is run inside its own 1.cleandata directory.

Step 1: Clean RNA-seq raw reads¶

mkdir -p ./sce/3.rna-seq/1.cleandata/
cd ./sce/3.rna-seq/1.cleandata/

for fq in ../../2.rawdata/rna-seq/*fastq.gz
do
  cutadapt --match-read-wildcards \
    -a AGATCGGAAGAGCGTCGTGTAGGGAAAGAGTGTAGATCTCGGTGGTCGC \
    -m 25 \
    -O 6 \
    -j 12 \
    -o $(basename $fq fastq.gz)clean.fastq.gz \
    $fq \
    &>> $(basename $fq .fastq.gz)".cutadapt.log"
done

Step 2: Clean Ribo-seq raw reads¶

Ribo-seq cleaning is similar to RNA-seq cleaning, but the adapter is replaced with the 3' adapter (e.g. poly-A) used in Ribo-seq library construction.

mkdir -p ./sce/4.ribo-seq/1.cleandata/
cd ./sce/4.ribo-seq/1.cleandata/

for fq in ../../2.rawdata/ribo-seq/*fastq.gz
do
  cutadapt --match-read-wildcards \
    -a AAAAAAAA \
    -m 25 \
    -O 6 \
    -j 10 \
    -o $(basename $fq fastq.gz)clean.fastq.gz \
    $fq \
    &>> $(basename $fq .fastq.gz)".cutadapt.log"
done

Parameter explanation¶

Parameter	Meaning
`--match-read-wildcards`	allow wildcard matching in reads
`-a`	adapter sequence
`-m`	minimum retained read length
`-O`	minimum adapter overlap length
`-j`	threads
`-o`	output FASTQ

Outputs¶

File	Description
`*.clean.fastq.gz`	cleaned FASTQ
`*.cutadapt.log`	adapter trimming statistics

QC interpretation¶

Check:

percentage of reads with adapters
retained read number
post-trimming read length distribution
Ribo-seq dominant RPF length range
RNA-seq read length consistency