4.5.2 Enzymatic bias

Purpose

rpf_Digest detects 5' and 3' end sequence preferences around read ends. These patterns can reflect nuclease digestion, ligation, or library-construction bias.

This step is useful because strong end-sequence bias can affect local RPF density and may mimic codon-level pausing signals.

The digestion-bias workflow is divided into two steps:

Step 1: Run rpf_Digest for each sample
Step 2: Merge digestion results

Input files

Input	Description
BAM file	Filtered BAM file generated by rpf_Check (Step 1 of 4.5.1), such as *.bam under 01.qc/.
Transcript annotation	Normalized transcript annotation file, usually gene.norm.txt generated by rpf_Reference.
Transcript FASTA	Normalized transcript sequence file, usually gene.norm.rna.fa generated by rpf_Reference.
Read length range	RPF read length range to retain (-m / -M). For Ribo-seq, this is often around 27--33 nt.

Step 1: Run rpf_Digest

1.1 Parameters

Parameter	Required	Description
-t / --transcript	Yes	Input transcript annotation file generated by rpf_Reference.
-s / --sequence	Yes	Input transcript sequence file in FASTA format.
-b / --bam	Yes	Input BAM alignment file.
-o / --output	Yes	Output file prefix.
-l / --longest	No	Only retain the transcript with the longest CDS per gene. Disabled by default.
-m / --min	No	Minimum read length retained. Default is 20 nt.
-M / --max	No	Maximum read length retained. Default is 100 nt.
--scale	No	Scale the digestion-site motif matrix and generate the scaled digestion-site plot. Disabled by default.
--thread	No	Number of worker processes for indexed BAM input. Default is 1.

1.2 Example

cd ./sce/4.ribo-seq/02.digestion/

for bam in ../01.qc/*.bam
do
    prefix_name=$(basename ${bam} .bam)

    rpf_Digest \
        -b ${bam} \
        -t ../../../1.reference/norm/gene.norm.txt \
        -s ../../../1.reference/norm/gene.norm.rna.fa \
        -o ${prefix_name} \
        -m 27 \
        -M 33 \
        --thread 8 \
        --scale \
        &> ${prefix_name}.log
done

1.3 Output

Output	Description
<prefix>_digestion_sites.txt	5' and 3' end frame counts by RPF length (5end_frame1-3, 3end_frame1-3).
<prefix>_digestion_sites_plot.pdf	Heatmap of 5' and 3' end frame distribution across RPF lengths.
<prefix>_scaled_digestion_sites_plot.pdf	Scaled digestion-site plot, generated when --scale is used.
<prefix>_5end_counts.txt	5' end nucleotide count table.
<prefix>_5end_pwm.txt	5' end nucleotide PWM table.
<prefix>_5end_seqlogo2.pdf	5' end sequence logo.
<prefix>_3end_counts.txt	3' end nucleotide count table.
<prefix>_3end_pwm.txt	3' end nucleotide PWM table.
<prefix>_3end_seqlogo2.pdf	3' end sequence logo.
<prefix>.log	Running log if redirected by the user.

Example output figures

The sequence logos show the nucleotide preference around the 5' and 3' read ends:

Step 2: Merge digestion results

merge_digestion merges the 5' or 3' digestion PWM files from multiple samples into a single table.

2.1 Parameters

Parameter	Required	Description
-l / --list	Yes	Input digestion PWM files. The file names must end with _5end_pwm.txt or _3end_pwm.txt, otherwise the command exits with an error. Multiple files can be provided.
-o	Yes	Output prefix. The output table is <prefix>_reads_digestion.txt.

2.2 Example

merge_digestion \
    -l *_5end_pwm.txt *_3end_pwm.txt \
    -o RIBO

2.3 Output

File	Description
RIBO_reads_digestion.txt	Merged digestion motif table across samples.
RIBO_5end_seqlogo.pdf / RIBO_5end_seqlogo.png	Merged 5' end sequence logos across samples (one row per sample).
RIBO_3end_seqlogo.pdf / RIBO_3end_seqlogo.png	Merged 3' end sequence logos across samples (one row per sample).

Example output figures

The merged sequence logos show the nucleotide preference around the 5' and 3' read ends across all samples (one row per sample):

Notes

• Strong nucleotide preference around read ends suggests digestion or ligation bias.
• Check both 5' and 3' end patterns before interpreting codon-level pausing signals.
• Use the same read-length range as downstream Ribo-seq density generation whenever possible.
• The -l input files of merge_digestion must end with _5end_pwm.txt or _3end_pwm.txt.