search

morrislab / qapa

RNA-seq Quantification of Alternative Polyadenylation
GNU General Public License v3.0
42 stars 10 forks source link

Can i use QAPA to quantify APA usage based on long reads transciptome data (nanopore or pacbio single molecule sequencing)? #46

Open operoncao opened 2 years ago

MustafaElshani commented 2 years ago

I'm also interested in this question regarding using APA with nanopore reads?

kcha commented 2 years ago

I don't have any experience with nanopore reads to date. My guess is if you are able to quantify the 3' UTRs using Salmon then it should work with QAPA.

MustafaElshani commented 2 years ago

Thank you very getting back, I tried salmon quant on long reads, the mapping rate is very low around 1.2% .

I think this is salmon issue, I will try other ways

MustafaElshani commented 2 years ago

Hi @kcha I think I have manged to get QAPA to work with nanopore, I'm trying to validate it somehow.

Needed to go back a few steps in the pipeline, default salmon mapping-mode gave very low mapping rate so I had to go back to alignment-based mode, but to do this I had to align long reads to the output_sequences.fa generated by qapa fasta. This was followed by aligning the reads using minimap2 this was done by the following commands

minimap2 -ax splice -t 40 output_sequences.fa NanoporeSample1.fastq | samtools view -Sb > QAPA_alignments/NanoporeSample1.bam

followed by some samtooling just incase

samtools sort -@ 40 -T tmp -o QAPA_alignments/NanoporeSample1_sorted.bam QAPA_alignments/NanoporeSample1.bam && samtools index QAPA_alignments/NanoporeSample1_sorted.bam

Salmon quant was then used

salmon quant --ont -p 40 -t output_sequences.fa -l U -a QAPA_alignments/NanoporeSample1_sorted.bam -o salmon_QAPA_Nanopore/NanoporeSample1_salmon

This was finilised with qapa quant qapa quant --db ensembl_identifiers.txt salmon_QAPA_Nanopore/*/quant.sf > pau_results_Nanopore.txt

Now i'm trying to validate if this has worked just need to understand some of the Scripting for visualisation of the data as in the question here https://github.com/morrislab/qapa/issues/38