DE seq analysis using transcript expression data

Question

[shivangi.agarwal800]

Hi

How to perform DE seq analysis using transcript expression data?

Regards

[Nitin Narwade]

first of all, you should have a raw read count table for each of your samples i.e. control as well as treatment. To prepare raw read count table you have to map your raw reads on the reference genome. If your doing Denovo assembly (i.e. Reference genome is not available for an organism) then you have to map all raw reads from each of your samples on Denovo assembled transcriptome (Denovo assembly can build using tools like Trinity) using read aligner like BWA, Bowtie, Bowtie2 etc.

And in case if you are having reference genome available for your organism then you can directly get the alignment file while running your pipeline.

Once you got the alignment file(let's say .bam file; here you have to convert primary alignment file i.e. .sam into .bam file using samtools view utility in response to reducing the file size for efficient file handling and space management.) further using samtools idxstat you can get the raw read count using alignment BAM file.

Further such raw read count table(s) you can use to find out DE using tools/packages such as: cuffdiff, ballgown, DESeq, DESeq2 , edgeR etc.

OR

As doctor.dee00 suggested you can use that methods too.

[lakhujanivijay]

further using samtools idxstat you can get the raw read count using alignment BAM file

samtools idxstat as the name suggest provides the index statistics. Be aware of it's limitations as it can loose a lot of useful information regarding RNA-seq reads mapped onto genome for example what fraction of the read mapped? Whether the read mapped onto multiple regions / genes/ exons and how much overlap. It's better too stick to standard tools like HT-seq.

Further, idxstats also provides count of unmapped reads which is confusing at first. For a unmapped read, it will assign the coordinates of it's mate which got mapped somewhere.

Answer 1

[Dattatray Mongad]

The best reference for this would be official documentation of (DESeq2). Or you can refer to following discussions. Transcript to gene level count for DEseq(2) use- Salmon/Sailfish/Kallisto etc. DESeq2 transcript level If you are looking for script you can find it here