herber4 / Quant_Seq_DaPars2

Analysis of alternative polyadenylation from QuantSeq Rev data with DaPars2

Differential APA analysis from Quant Seq Rev data using DaPars2

This is an imperfect protocol and still under development

Align reads with STAR, soft alignment constraints. Convert sam to bam, sort, index and gather coverage info

ml star/2.7.10a

for i in *_R1_001.fastq.gz; do name=$(basename ${i} _R1_001.fastq.gz);

STAR --runThreadN 32 \
--readFilesCommand zcat --genomeDir /data2/lackey_lab/austin/dbs/star \
--readFilesIn ${name}_R1_001.fastq.gz ${name}_R2_001.fastq.gz \
--outFilterType BySJout --outFilterMultimapNmax 200 \
--outFilterMatchNminOverLread 0.33 --outFilterScoreMinOverLread 0.33 \
--alignSJoverhangMin 8 --alignSJDBoverhangMin 1 --outFilterMismatchNmax 999 \
--outFilterMismatchNoverLmax 0.6 --alignIntronMin 20 --alignIntronMax 1000000 --alignMatesGapMax 1000000 \
--limitOutSJcollapsed 5000000 \
--quantMode TranscriptomeSAM \
--outSAMattributes NH HI NM MD --outSAMtype BAM Unsorted \
--limitBAMsortRAM 2000000000 \
--outFileNamePrefix ${name}_ ;
done

module load samtools/1.10
for B in *_Aligned.out.bam; do
    N=$(basename $B _Aligned.out.bam) ;
    samtools view --threads 8 -bS $B | samtools sort -O BAM -o ${N}.sorted.bam ;
done

for b in *.sorted.bam ; do
	samtools index $b ;
done

for b in *.sorted.bam; do
    N=$(basename $B .sorted.bam) ;
    samtools coverage -o ${N}.txt $b ;
done

Generate genome coverage bedgraphs from bam files

for b in *.sorted.bam; do
	N=$(basename $b .sorted.bam) ;
	bedtools genomecov -ibam $b -bga -split -trackline > $N.bedgraph ;
done

Generate samtools coverages to coverage input in DaPars2

module load samtools/1.10

for b in *.sorted.bam; do
    N=$(basename $b .sorted.bam) ;
    samtools coverage -o ${N}.txt $b ;
done

run the reduce_chrs.sh script to get chromosomes from samtools coverage output

#!/bin/bash

for t in *.txt ; do
	N=$(basename $t .txt) ;
	head -n 26 $t > $N.reduced.txt ;
done

run this script to get coverage sums from each sample

#!/bin/bash

for t in *.txt ; do
	N=$(basename $t .txt) ;
	awk 'NR > 1 { sum += $4 } END { print sum }' $t > $N.counts ;
done
cat *.counts > merged.counts

Running Dapars2

• examples of files you need are in the github_dapars2 dir
• make sure all bedgrpahs and necessary files are in the same directory to run
• items needed to run: bedgraph for each sample chrList.txt dapars.config sample_coverages.txt RefSeq_hg38_3UTR_annotation.bed

conda activate dapars

python3 ../DaPars2/src/DaPars2_Multi_Sample_Multi_Chr.py dapars.config chrList.txt

Moving raw dapars2 for statistical analysis in R

You need to move all chromosome output into one dir, as well as create a meta data file like the sample_meta.txt, you can do this in excel

mkdir output
cp */*.txt output/